Diagnostic Performance of Photoplethysmography (PPG) for Early Vascular Compromise in a Customisable In Vitro Flap Model

Diagnostic Performance of Photoplethysmography (PPG) for Early Vascular Compromise in a Customisable In Vitro Flap Model | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Diagnostic Performance of Photoplethysmography (PPG) for Early Vascular Compromise in a Customisable In Vitro Flap Model Hiroki Kodama, James May, Dariush Nikkhah, Panicos A Kyriacou This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8080454/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background Free flap reconstruction is a cornerstone of plastic surgery, yet vascular thrombosis, a blockage of the supporting blood vessels,(3%–5% incidence) requires early detection for salvage. Traditional clinical assessment is subjective and labour-intensive, driving the need for reliable monitoring. Photoplethysmography (PPG), a non-invasive, continuous optical sensing technology monitoring vascular volume, is highly suited for this purpose. Methods This study investigated PPG's diagnostic capability for early vascular changes utilising a custom silicone phantom as a test-bed. Vessels were fabricated to mimic human arterial and venous mechanics (compliance, ΔCSA) and embedded at depths of 3,9,15, and 21 mm. A perfusion system simulated normal, early ischemic (low arterial pressure), and early congested (high venous pressure) states. Signals were analysed using 34 pulse-wave features. Results Analysis of the diagnostic performance (AUC > 0.70) showed that PPG (IR) could successfully diagnose ischemia up to a depth of 21 mm and congestion up to 15 mm. The system was able to successfully distinguish between ischemia and congestion at all depths. Conclusion This research confirms that specific PPG features are discernible indicators of early blood flow changes. The high diagnostic accuracy achieved in this in vitro model establishes PPG as a robust platform for developing future continuous and objective free flap monitoring systems. Health sciences/Diseases Physical sciences/Engineering Health sciences/Health care Health sciences/Medical research Free Flap Photoplethysmography (PPG) Phantom study Plastic Surgery Biomedical Engineering Wearable Device Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 1. Introduction Free flaps are tissue units transferred from a donor to a recipient site while maintaining their original blood supply via microvascular anastomosis (connection between two blood vessels). Free flap reconstruction represents a cornerstone method for managing large and complex defects, offering superior functional and aesthetic outcomes in diverse anatomical locations, including the head and neck 1 , breast 2 , and extremities 3 . The procedure's growing importance is evidenced by national data; for example, a 2020 report from the United States indicated that the number of free flap procedures performed for breast reconstruction reached 23,324, representing a 75% increase over the preceding two decades 4 . Despite the substantial benefits of microvascular tissue transplantation, blood flow deficits remain a significant concern, occurring in approximately 3% to 5% of cases 1 , 5 . Vascular thrombosis is the principal risk factor, typically manifesting within the first 48 hours post-surgery 6 – 8 . A severe consequence of delayed diagnosis is a significantly reduced probability of free flap salvage. Furthermore, the need for a second flap to address failure is reported to incur an additional cost of $ 174,000 9 . Crucially, historical studies report that prompt salvage surgeries initiated within the first 24 hours of vascular compromise successfully save 70% to 80% of affected flaps 5 , 10 – 12 . The traditional method of clinical assessment remains the established standard for evaluating free flaps and achieving early detection of compromise. This evaluation includes measurements of skin colour, capillary refill time, tissue fullness (turgour) caused by fluid build up, perceptible by the physical examination of how hard or soft the tissue is, temperature, and pinprick response. However, these observations are labour-intensive, as they must be performed manually and frequently 13 . Furthermore, these characteristics are qualitatively and subjectively monitored, largely depending on the expertise of the individual physician or nurse. Because different personnel often monitor the flap, inconsistencies may arise in these observations, risking the oversight of subtle changes 13 . Various monitoring devices are currently available on the market. Pinpoint measurement, devices such as Near-Infrared Spectroscopy (NIRS) 14 , Laser Doppler Flowmetry (LDF) 15 , implantable Doppler probes 16 , flow couplers 17 , and microdialysis 18 are utilised. For image-based assessment, techniques such as Indocyanine Green Fluorescence Angiography 19 and thermography 20 are available. Each of these technologies possesses inherent strengths and limitations, and a universal consensus regarding their optimal clinical use does not currently exist 21 . Photoplethysmography (PPG), alongside NIRS, is an optical sensing technology that allows for non-invasive and continuous monitoring. While NIRS measures tissue-level oxygenation, PPG monitors changes in vascular volume. A PPG sensor primarily consists of two components: a light source, typically a Light Emitting Diode (LED), and a photodetector (such as a photodiode or phototransistor) that measures the amount of received light 22 . Depending on the body site where the sensor is placed, it operates in either transmission or reflectance mode. In transmission mode, red and infrared light from the LED passes through a body part and is detected by a photodetector positioned on the opposite side. In reflectance mode, the photodetector measures the amount of light reflected back from the tissue 22 . Due to its small size and low cost, PPG has long been used as the fundamental technology for pulse oximetry in clinical settings. Recent technological advancements have led to this technique being widely incorporated into wearable devices such as smartwatches and rings, and its clinical significance is rapidly increasing 23 . This sensor technology enables continuous data acquisition, making it possible to detect abnormalities in a timely manner that cannot be achieved with static images. However, direct evidence regarding PPG's specific diagnostic capability in free flap monitoring remains limited 24 , 25 . An ideal PPG signal is characterised by a systolic peak and a diastolic peak, separated by the dicrotic notch. However, in practice, the diastolic peak is frequently absent, particularly in elderly patients 26 , 27 . In this study, we adopted a feature extraction algorithm based solely on the systolic peak from the PPG waveform, and the features were used to identify the correlation with early ischemia and congestion. Furthermore, the features analysed rely on the original filtered PPG waveform, rather than its derivative. Consequently, this correlation is applicable even when the second derivative reference points used in existing pulse wave analysis methods are unavailable 28 . This paper aims to investigate the diagnostic capability of PPG for early blood flow changes in free flaps by evaluating PPG signals acquired from an in vitro vascular system. By demonstrating the utility of various pulse wave characteristics, this work is expected to inform future monitoring technologies. The research is predicated on the hypothesis that specific morphological features of the pulse wave serve as discernible indicators of early blood flow changes. 2. Methods 2.1. Phantom Creation and Experimental Setup The experimental system was built using established Zen PPG hardware 29 . To simulate free flap vasculature, custom-made silicone vessels were fabricated according to previous methods 30 , 31 , accurately mimicking the ΔCSA (Change in Cross-Sectional Area) and compliance reported in the literature 32 . The general experimental circuit for PPG validation was modified to specifically model flap circulation, incorporating the fabricated vessels at varying depths within a tissue phantom. Intravascular pressure was precisely controlled to simulate the small artery and vein under specific conditions, including normal flow, early ischemia, and early congestion. This setup allowed for an in vitro assessment by creating precise and reproducible flow abnormalities. 2.1.1. Phantom Creation The phantom creation was based on previous research conducted in our laboratory 30 , 31 , 33 . The mechanical properties of the vessels were reproduced by mixing equal parts of Gel-00 (Polytek Development Corp., Easton, PA, USA) Part A and Part B with appropriate ratios of deadener (D) and hardener (H), along with a 2% retarder 31 . The wall thickness (d) and outer diameter (OD) of the fabricated vessels were measured. Based on the degree of expansion in response to the appropriate internal pressure, ΔCSA and compliance were calculated to determine the optimal blend ratio for mimicking human tissue. Four separate vascular tissue phantoms were fabricated by embedding the vessels (artery and vein) within a surrounding silicone matrix. This process, which followed the previously established casting procedure 30 , 31 , positioned the vessels at depths of 3, 9, 15, and 21 mm from the phantom's surface (Fig. 1 (a)). 2.1.2. In Vitro Cardiovascular System Configuration The fabricated tissue phantom was incorporated into an in vitro vascular system, replicating the systemic circulation system of the human body. The phantom was positioned within a dedicated sensor casing to ensure consistent monitoring. A reflectance PPG sensor, featuring Red (660 nm) and Infrared (940 nm) LEDs, was placed on the underside of the phantom. This sensor was connected to a custom dual-channel PPG acquisition system, developed by the Research Centre for Biomedical Engineering (RCBE) group at City St George’s, University of London 34 , 35 , which operates at a sampling rate of 2000 Hz. The signals were displayed and recorded using LabVIEW (version 2023 Q1, National Instruments, Austin, Texas). To simulate blood flow, a liquid composed of deionized water and methylene blue powder (Thermo Fisher Scientific, UK) 36 was circulated throughout the system using a pulsating pump (PD-1100, BDC Laboratories, Wheat Ridge, CO) operating at 60–65 bpm. The phantom vein was assumed to have minimal pulsation and was therefore connected to a closed circuit with adjustable internal pressure (Fig. 1 (b)). Based on previous studies, arterial internal pressure was set to 40–80 mmHg for the normal and congested state and 25–50 mmHg for the ischemic state 37 , 38 . The adjustment of internal pressure was done by adjusting a clamp within the circuit. The venous internal pressure, on the other hand, was set to 10 mmHg for the normal and ischemic states and 30 mmHg for the congested state, which was adjusted by injecting fluid via a syringe through a valve (Fig. 1 (b)). 2.2. Signal Processing and Statistical Analysis Figure 2 illustrates the flow of procedures from signal processing through statistical analysis. The acquired signals were assessed for quality by two independent experts, using the following criteria: 'Excellent' (clear systolic and diastolic waves), 'Acceptable' (heart rate detectable), and 'Poor' (heart rate is difficult to detect). The signals were further evaluated using the Perfusion Index (AC/DC Ratio) before proceeding to the analysis stage. PPG signal preprocessing was performed following a standard signal processing pipeline consisting of four sequential stages: (1) bandpass filtering, (2) artifact detection, (3) artifact removal, and (4) baseline correction. To preserve the primary physiological components of PPG signals (cardiac-related frequency components) while removing high-frequency noise and low-frequency drift, a bandpass filter with cutoff frequencies of 0.5–10 Hz was applied 39 . A 4th-order Chebyshev Type II filter was employed for filter design 40 . The stopband attenuation was set to 40 dB, and zero-phase filtering (forward-backward filtering) was implemented to avoid phase distortion 41 . Following filtering, artifacts were detected using the MAD (Median Absolute Deviation) method, a statistical outlier detection technique 42 . The modified Z-score was calculated as follows… $$\:Modified\:Z-score=0.6745\times\:\frac{(Xi-median)}{MAD}$$ …where MAD = median(|Xi - median|), and 0.6745 is the conversion factor from MAD to standard deviation under the assumption of normal distribution. Data points with an absolute modified Z-score exceeding 3.5 were identified as artifacts 43 . Detected artifact points were replaced using linear interpolation 44 . Specifically, values at artifact positions were estimated by linear interpolation using valid signal points that were not identified as artifacts. In this study, baseline estimation and correction were performed using the cubic spline interpolation method 45 , 46 . First, the baseline component was estimated as the lower envelope of the signal using a moving minimum filter with a window size corresponding to approximately one cardiac cycle (0.8 s, 1600 sample points). Next, a cubic spline function was fitted to the estimated baseline points with a smoothing parameter of s = 0.01 × N (where N is the number of sample points) to obtain a smoothed baseline curve. Finally, the baseline-corrected PPG signal was obtained by subtracting this baseline curve from the original signal. Through these preprocessing steps, high-quality signals with reduced noise and baseline variations were obtained for both Red and IR channel PPG signals. PPG features were extracted from the pulse waves (Fig. 3 , Table 1 ) 47 . Table 1 Table of PPG features Feature (abbreviation) Description Area Under Curve (AUC) Area of pulse Systolic Area Under Curve (S-AUC) Area of the systolic phase of the pulse Diastolic Area Under Curve (D-AUC) Area of the diastolic phase of the pulse Area Under Curve Ratio (AUC Ratio) Ratio between systolic and diastolic areas Rise Time Time taken for signal to rise from trough to peak Decay Time Time taken for signal to fall from peak to trough Rise-Decay Time Ratio Ratio of decay time to rise time Amplitude Median amplitude Upslope Length Length of tangent from trough to peak Downslope Length Length of tangent from peak to trough Upslope Gradient of pulse on rising edge Downslope Gradient of pulse on falling edge Onset-End Slope Rate of change of the pulse over the entire pulse length Slope Ratio Rising slope gradient (upslope) compared to falling slope gradient (downslope) Length-Height Ratio Ratio of the height of a pulse compared to its width Slope Length Ratio Ratio of the upslope length of a pulse compared to its downslope length Upslope Length Ratio Proportion of the total pulse length that is made up of the rising phase Downslope Length Ratio Proportion of the total pulse length that is made up of the falling phase Start Datum Area Area between PPG signal and the tangent from trough to peak (systolic phase) End Datum Area Area between PPG signal and the tangent from peak to trough (diastolic phase) Datum Area Ratio Ratio between systolic and diastolic datum areas Max Start Datum Difference Maximum length between PPG signal and peak-trough tangent (systolic phase) Max End Datum Difference Maximum length between PPG signal and trough-peak tangent (diastolic phase) Median Start Datum Difference Median length between PPG signal and peak-trough tangent (systolic phase) Median End Datum Difference Median length between PPG signal and trough-peak tangent (diastolic phase) Pulse Width Width of the entire pulse of one cycle (measured at 50% of height) Systolic Width Width of the systolic phase of the pulse (measured at 50% of height) Diastolic Width Width of the diastolic phase of the pulse (measured at 50% of height) Width Ratio Ratio between the systolic and diastolic widths (measured at 50% of height) Variance Deviation of signal from its mean value Skewness Degree of symmetry of a pulse Kurtosis Degree of sharpness of a pulse Signal-to-Noise Ratio (SNR) Level of PPG signal compared to noise Zero-Crossing Rate (ZCR) Number of times per second that the PPG signal crosses zero Peak-To-Instantaneous Ratio (PIR) Degree of change in amplitude over time The diagnostic performance for differentiating the Normal, early Ischemic, and early Congested states was evaluated using a comprehensive feature set. Receiver Operating Characteristic (ROC) analysis was performed for each pairwise comparison, wavelength, and depth to complement group-level p-values with a measure of classification ability. Diagnostic accuracy was quantified by generating these ROC curves based on sensitivity and false-positive rates, with the Area Under the Curve (AUC) serving as the final performance metric to quantify how well the normalised amplitude could distinguish between two states. To ensure the robustness of the AUC estimates, 95% confidence intervals (CIs) were derived using a bootstrap resampling method (N = 200 resamplings). Furthermore, the optimal threshold (cutoff) for each feature and each pair was identified by maximising Youden's J statistic (sensitivity + specificity − 1), and the corresponding sensitivity and specificity were recorded. Features were defined as possessing diagnostic performance if their Area Under the Curve (AUC) exceeded 0.7 48 . All statistical analyses and visualizations were performed in Python 3.13.0. 3. Results 3.1. Phantom Fabrication and Mechanical Characterization The change inΔCSA and compliance were calculated as shown in Fig. 4 (a). The blend ratio for the arterial phantom was adjusted to approximate a ΔCSA range of 23.5%–25.9% and a compliance range of 1.3–3.52×10 − 4 cm 2 /mmHg across pressure changes of 40–80 mmHg and 60–120 mmHg 32 . The final ratio of the Hardener was determined to be 0.6, resulting in an initial internal diameter (ID 0 ) of 2.72 mm(Fig. 4 (b)). Similarly, the blend ratio for the venous phantom was adjusted to approximate a ΔCSA range of 27.2%–41.8% and a compliance range of 5.6–17.4×10 − 4 cm 2 /mmHg across pressure changes of 10–30 mmHg 32 . The final ratio of the Deadener (D) was determined to be 0.4, resulting in an initial ID 0 of 3.19 mm (Fig. 4 (c)). These artery and vein phantoms were then embedded 3, 9, 15, 21 mm deep within a subcutaneous tissue phantom (ratio of D was 1) created based on a previous report 49 , 30 , 31 . 3.2. Signal Acquisition Across each wavelength, depth, and condition, the appropriate internal pressures were successfully replicated, and stable waveforms were acquired (Fig. 5 ). 3.3. Signal Quality Assessment Signal quality, assessed independently by two PhD students, was deemed 'Excellent' (clear systolic and diastolic waves) or 'Acceptable' (heart rate detectable) for the Red signal up to a depth of 9 mm and for all IR signals 50 . Furthermore, the calculated Perfusion Index (PI: AC/DC Ratio) was 0.13%–0.43% (Table 2 ), which falls within the previously reported range of 0.12%–0.76% 51 . Table 2 Perfusion Index (PI) at each depth Depth State Channel Quality PI (%) 3 Normal Red Excellent 0.35 IR Excellent 0.43 Ischaemia Red Excellent 0.22 IR Excellent 0.39 Congestion Red Excellent 0.43 IR Excellent 0.38 9 Normal Red Acceptable 0.17 IR Excellent 0.21 Ischaemia Red Acceptable 0.14 IR Excellent 0.16 Congestion Red Acceptable 0.18 IR Excellent 0.22 15 Normal Red Unfit 0.16 IR Excellent 0.16 Ischaemia Red Unfit 0.15 IR Excellent 0.13 Congestion Red Unfit 0.16 IR Excellent 0.17 21 Normal Red Unfit 0.18 IR Excellent 0.16 Ischaemia Red Unfit 0.17 IR Acceptable 0.18 Congestion Red Unfit 0.19 IR Excellent 0.14 3.4. Diagnostic Ability Diagnostic performance was defined by an Area Under the Curve (AUC) exceeding 0.7. These diagnostically useful features are presented in Fig. 6 , segmented by pairwise condition and depth. Amplitude proved to be the most reliable feature for distinguishing between normal and abnormal perfusion states. While detecting early congestion at the deepest tissue levels presented a challenge, various other extracted features collectively enabled discrimination for all remaining diagnostic pairs. 4. Discussion The required frequency for free flap assessment, with frequency as high as every hour to every 30 minutes 52 , has historically placed a substantial burden on healthcare systems, primarily relying on in-house resident physicians 13 . In the current environment, which prioritises strict limitations on working hours, this manual monitoring approach is reaching its practical limit 13 , creating a strong desire for reliable monitoring devices. Despite this, a culture of reliance on manual clinical checks persists. Creech and Miller 53 outlined the essential criteria for an effective flap perfusion monitoring system, requiring it to be accurate, reliable, applicable to all free flaps, easy for medical personnel to use, safe for both the patient and the flap, and capable of early detection. Swartz 54 further argued that an ideal monitoring method should provide continuous record patency and flow at the anastomosis, distinguish between arterial and venous compromise, and be applicable to both cutaneous and buried flaps. A consistent feature across all free flap procedures, regardless of the presence of a skin paddle (the portion of skin included with the flap from the donor site), is the subcutaneous location of the pedicle (the nourishing vessels) and the relatively limited variation in its vessel caliber 55 (diameter of blood vessel). Our experimental design focused on developing a system that non-invasively accesses this pedicle, allowing for continuous measurement, discrimination between arterial and venous issues, and usability in buried flaps (transferred without a skin paddle, such as a muscle or bone flap, which is then covered by the native skin and is not visible on the surface). The distinctive temporal window of blood flow compromise,with 80% of thrombosis occurring within two days post-surgery, 90% of arterial thrombosis within 24 hours 8 , and salvage rates dropping significantly after five hours 7 ,is rarely seen in other disease groups. Developing a dedicated device focused solely on these tight parameters is commercially challenging due to the small market size, thereby stifling innovation. Consequently, achieving "commercially available" necessitates the application of established technological principles. In this context, PPG represents one of the most suitable technologies. In this study, silicone vascular phantoms were first fabricated. To evaluate their mechanical properties, Young’s modulus measurement and internal pressure testing were commonly considered. Internal pressure testing was chosen because it directly assessed compliance and the pressure–diameter relationship under conditions similar to physiological blood flow 58 . In the internal pressure testing performed in this study, circumferential compliance was calculated, while longitudinal strain was not considered. This was because preliminary tests showed that when an internal pressure of 30 mmHg was applied to a soft silicone tube with a deadener ratio of 0.6, the axial elongation was only approximately 1%. The relationship between circumferential strain (εθ) and longitudinal strain (εz) can be expressed using Poisson’s ratio (ν) as follows 59 : $$\:\epsilon\:\theta\:=\:\epsilon\:z\:\bullet\:2\:\bullet\:\frac{1-v/2}{1-2v}$$ Since ν is approximately 0.49 for soft silicone, εθ was calculated to be about 76 times greater than εz. Therefore, longitudinal strain was considered negligible. The pressure range of 40–80 mmHg was set to simulate the internal thoracic artery, and 60–120 mmHg was used to simulate the radial artery. As expected, increasing the deadener ratio increased both the change in cross-sectional area (ΔCSA) and compliance, whereas increasing the hardener ratio had the opposite effect. The arterial model showed the best agreement with physiological artery behaviour at a hardener ratio of H = 0.6, where ΔCSA was 6.62–8.09 mm² (+ 22.5%) at 40–80 mmHg and 7.34–9.14 mm² (+ 24.4%) at 60–120 mmHg, which falls within reported values for the radial artery (3.4–4.2 mm², + 23.5%) and brachial artery (8.1–10.2 mm², + 25.9%) 32 . The compliance of the arterial model (2.99 × 10⁻⁴ cm²/mmHg) was also comparable to reported physiological values for the radial (1.3 × 10⁻⁴ cm²/mmHg) and brachial arteries (3.52 × 10⁻⁴ cm²/mmHg) 32 . In terms of scale, the cross-sectional area also matched the reported range for the deep inferior epigastric artery (3.14–9.62 mm²) 60 . Similarly, the venous model best reproduced physiological venous mechanics at a deadener ratio of D = 0.4, where ΔCSA was 9.36–12.5 mm² (+ 33.6%) at 10–30 mmHg, which lies between reported values for the cephalic vein (5.5–7.8 mm², + 41.8%) and basilic vein (16.9–21.5 mm², + 27.2%) 32 . The compliance of the venous model (15.7 × 10⁻⁴ cm²/mmHg) was also consistent with reported venous values for the cephalic (5.6 × 10⁻⁴ cm²/mmHg) and basilic veins (17.4 × 10⁻⁴ cm²/mmHg) 32 . In terms of scale, the cross-sectional area was consistent with the reported range for the deep inferior epigastric vein (0.20–14.50 mm²) 60 . These arteries and veins were embedded in a tissue phantom whose silicone mixture ratio (A:B:D = 1:1:1) was formulated to achieve a hardness corresponding to that of the finger and thigh 61 , 62 . The internal pressure settings used for simulating the initial ischemic and congested pedicle states were established based on consultation with researchers who have performed continuous arterial and venous catheter-based pressure monitoring in over 200 free flap cases 38 . These settings primarily model pressures relevant to breast reconstruction; more pronounced pressure variations would be anticipated in reconstructions of the head and neck or extremities. Our preprocessing methodology utilised literature-validated standard approaches, including the 4th-order Chebyshev Type II bandpass filter 40 , MAD-based artifact detection 42 , and cubic spline interpolation for baseline correction 44 . However, we acknowledge that no standardised preprocessing pipeline for PPG signals currently exists, with optimal cutoff frequencies varying significantly across studies 63 . While the literature notes that optimal filtering parameters vary across in vivo subjects and conditions 63 , our approach employs a fixed bandpass filter (0.5–10 Hz) 39 designed to cover the physiologically relevant heart rate range. Given the highly controlled, stationary nature of our in vitro phantom and perfusion system, the established, fixed-parameter preprocessing pipeline ensures consistent and reliable signal conditioning across all experimental states. Despite the validity of our literature-supported fixed approach, this limits our study. Future research should implement signal-specific adaptive optimisation. Such personalised filtering would offer substantial benefits for achieving higher precision in challenging measurement scenarios, such as the deep tissue measurements (21 mm) and pathological states (ischemia, congestion) investigated in this study. The evaluation of the Receiver Operating Characteristic (ROC) curves provides direct insight into the diagnostic capability of PPG features for distinguishing between vascular states. The results indicate that PPG features can separate free flap compromise based on distinct morphological characteristics. In the discrimination of Normal vs. Ischaemia, which represents a flow-occlusion anomaly marked by a drop in arterial pressure (AP), the IR Amplitude and IR Area features (e.g., IR_S_AUC, IR_D_AUC) achieved extremely high AUC values, ranging from 0.95 to 1.00. This observation demonstrates that Ischaemia directly causes a reduction in AP-dependent PPG amplitude and vascular volume. Specifically, the near-infrared channel accurately quantifies the absolute decrease in deep vessel blood volume, a primary indicator of ischemia, with near-perfect accuracy (AUC 1.00 at 15 mm). The high AUC values for diastolic area (D_AUC, 0.80–0.97) are consistent with the corresponding reduction in diastolic vascular volume due to reduced arterial inflow. Time-based features like IR_Systolic Width (0.90) and IR_RiseTime (0.90) were also effective but less discriminative than the amplitude/volume metrics. For Normal vs. Congestion, an anomaly marked by elevated venous pressure (VP) and restricted outflow, the results were similar: IR Amplitude (0.96–0.99) and Area features (0.87–0.99) showed strong discriminatory power. Congestion causes blood to pool in the veins, leading to an increased vascular baseline (DC component) and suppression of the arterial pulsation amplitude (AC component). This confirms that PPG accurately captures the physical damping of the pulse wave due to venous stasis. The high accuracy of time-based features, such as IR_FallTime (0.94 at 3 mm) and IR_SlopeRatio (0.93), also confirms that the physical waveform elongation caused by delayed outflow is a simultaneous and detectable morphological marker. The comparison of Ischaemia vs. Congestion proved to be the most complex, as absolute amplitude is not sufficient for distinction. Discrimination was primarily driven by time and gradient features. IR_FallTime (0.97 at 3 mm) was particularly effective, differentiating the conditions based on the contrast between the prolonged outflow period in Congestion and the potentially faster pulse decay observed during Ischaemia due to low AP. Furthermore, the UpslopeLength_norm feature achieved an AUC of 1.00 at 9 mm, suggesting a complete difference in the complexity of the systolic rise profile between the two abnormalities. The relative strength of the Red/IR_Amplitude_ratio (0.78 at 21 mm) in deep tissue suggests that the physical difference in penetration depth allows the two wavelengths to capture complementary information about blood flow changes at varying tissue layers, which aids in differential diagnosis at depth. In summary, PPG features effectively classify free flap vascular compromise from a morphological perspective. The distinction of Ischaemia/Congestion vs. Normal states is predominantly achieved through IR Amplitude and Area features (0.96–1.00), signifying that absolute vascular volume change is the core discriminator. In contrast, the vital discrimination between Ischaemia vs. Congestion relies on time and shape features (e.g., FallTime, Width Ratio), demonstrating that PPG accurately isolates the distinct physiological mechanisms of arterial pump failure versus venous outflow obstruction with high accuracy. This study employed phantom experiments, rather than human or animal clinical trials, to validate the theoretical diagnostic accuracy of Photoplethysmography (PPG) for early ischemia and early congestion in free flaps. Given that the incidence of vascular compromise in humans is low, ethical induction is difficult, and animal models present a scale mismatch, this research provides valuable insights. A limitation of this study is the technical inability to perfectly replicate the dermis and subcutaneous tissue overlying the pedicle. Changes in blood flow within these superficial tissues could influence the measured values. However, if measurements are taken via the flap's skin paddle, this change in measured value might actually accentuate the AUC for differentiation. Future human clinical trials are required to clarify this relationship. 5. Conclusion With recent advancements in PPG technology, an in vitro setting that can simulate normal, ischaemic, and congested states would be a beneficial tool for studies in free flap monitoring. The purposes of this study were twofold: 1) to develop a custom-made silicone phantom that simulates a free flap and to replicate ischaemic and congested states, and 2) to assess the capability to diagnose early flap disorder. PPG signals obtained from this model were recorded and analysed, allowing us to detect the morphological changes caused by each condition. A statistically significant change was detected, which demonstrates the potential of PPG technology for detecting flap blood flow impairment. Going forward, work should also aim to further improve the similarity of the phantom to human tissue by exploring different materials and generating signals that are even closer to those found in the human body. Furthermore, it is necessary to investigate more advanced signal feature extraction and highly accurate anomaly detection methods. Declarations Funding Declaration This research was supported by the Pump Priming Funds for consumables awarded by the British Association of Plastic, Reconstructive and Aesthetic Surgeons (BAPRAS). The authors would like to express their sincere gratitude to BAPRAS for the financial support that made this work possible. Author Contribution Hiroki Kodama: Conceptualization, Investigation, Formal analysis, Writing -original draft, Project administration, Methodology, Visualization, Data curationJames May: Conceptualization, Writing -reviewing and editingDariush Nikkhah: Writing -reviewing and editingPanicos Kyriacou: Writing -reviewing Acknowledgement I sincerely thank Dr. Saiga for his profound knowledge and expert guidance on setting the precise internal arterial and venous pressure parameters for the in vitro experiments, drawing from his extensive experience with over 200 cases of continuous catheter-based pressure monitoring in free flaps. I also wish to thank my laboratory colleagues for laying the foundational work of this research and my colleagues who assisted with the waveform quality assessment. Data Availability The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request. References Nakatsuka, T. et al. Analytic Review of 2372 Free Flap Transfers for Head and Neck Reconstruction Following Cancer Resection. J. reconstr. Microsurg . 19 (06), 363–368. 10.1055/s-2003-42630 (2003). Harashina, T., Imai, T., Nakajima, H. & Fujino, T. Breast reconstruction with microsurgical free composite tissue transplantation. Br. J. Plast. Surg. 33 (1), 30–37. 10.1016/0007-1226(80)90048-X (1980). Taylor, G. I., Miller, G. D. H., Ham, F. J. & THE FREE VASCULARIZED BONE GRAFT: A Clinical Extension of Microvascular Techniques. Plast. Reconstr. Surg. 55 (5), 533–544. 10.1097/00006534-197505000-00002 (1975). 2020 Plastic Surgery Statistics Report. American Society of Plastic Surgeons Accessed October 13. (2025). https://www.plasticsurgery.org/documents/news/statistics/2020/plastic-surgery-statistics-full-report-2020.pdf Caine, P., Jeevaratnam, J. A., Misky, A. & Nikkhah, D. Post-operative Monitoring. In: Core Techniques in Flap Reconstructive Microsurgery - A Stepwise Guide. Springer Nature Switzerland AG; :39–46. (2023). Tran, N. V., Buchel, E. W. & Convery, P. A. Microvascular Complications of DIEP Flaps. Plast. Reconstr. Surg. 119 (5), 1397–1405. 10.1097/01.prs.0000256045.71765.96 (2007). Bui, D. T. et al. Free Flap Reexploration: Indications, Treatment, and Outcomes in 1193 Free Flaps. Plast. Reconstr. Surg. 119 (7), 2092–2100. 10.1097/01.prs.0000260598.24376.e1 (2007). Kroll, S. S. et al. Timing of Pedicle Thrombosis and Flap Loss after Free-Tissue Transfer. Plast. Reconstr. Surg. 98 (7), 1230–1233. 10.1097/00006534-199612000-00017 (1996). Mericli, A. F. et al. A Prospective Clinical Trial Comparing Visible Light Spectroscopy to Handheld Doppler for Postoperative Free Tissue Transfer Monitoring. Plast. Reconstr. Surg. 140 (3), 604–613. 10.1097/PRS.0000000000003600 (2017). Disa, J. J., Cordeiro, P. G. & Hidalgo, D. A. Efficacy of conventional monitoring techniques in free tissue transfer: an 11-year experience in 750 consecutive cases. Plast. Reconstr. Surg. 104 (1), 97–101 (1999). Chen, K. T. et al. Timing of Presentation of the First Signs of Vascular Compromise Dictates the Salvage Outcome of Free Flap Transfers. Plast. Reconstr. Surg. 120 (1), 187–195. 10.1097/01.prs.0000264077.07779.50 (2007). Al-Dam, A. et al. Outcome of microvascular free flaps in a high-volume training centre. J. Cranio-Maxillofacial Surg. 42 (7), 1178–1183. 10.1016/j.jcms.2014.02.005 (2014). Patel, U. A. et al. Free Flap Reconstruction Monitoring Techniques and Frequency in the Era of Restricted Resident Work Hours. JAMA Otolaryngol. Head Neck Surg. 143 (8), 803. 10.1001/jamaoto.2017.0304 (2017). Kagaya, Y. & Miyamoto, S. A systematic review of near-infrared spectroscopy in flap monitoring: Current basic and clinical evidence and prospects. J. Plast. Reconstr. Aesthetic Surg. 71 (2), 246–257. 10.1016/j.bjps.2017.10.020 (2018). Kodama, H. et al. The future of free flap monitoring by laser continuous doppler flowmetry: A prospective assessment in consecutive 71 patients. JPRAS Open. 43 , 140–152. 10.1016/j.jpra.2024.11.004 (2025). Wu, K. G. et al. Implantable arterial doppler efficacy in free flap reconstruction: A systematic review and meta-analysis. Head & Neck . Published online August 29, :hed.27500. (2023). 10.1002/hed.27500 Colakoglu, S., Johnson, A., Anderson, J., Mathes, D. W. & Chong, T. W. Changes to Venous Flow Coupler Signal during DIEP Flap Inset Can Be Predictive of Poor Clinical Outcomes in Autologous Breast Reconstruction. J. Reconstr. Microsurg . 36 (6), 466–470. 10.1055/s-0040-1703014 (2020). Hwang, S. J. Diagnostic Accuracy of Microdialysis in Postoperative Flap Monitoring. J. Craniofac. Surg. 34 (1), 288–290. 10.1097/SCS.0000000000008878 (2023). Burnier, P., Niddam, J., Bosc, R., Hersant, B. & Meningaud, J. P. Indocyanine green applications in plastic surgery: A review of the literature. J. Plast. Reconstr. Aesthetic Surg. 70 (6), 814–827. 10.1016/j.bjps.2017.01.020 (2017). Hudson, T., Hogue, E., Mullner, D., Herrera, F. & Scomacao, I. The Utility of Smartphone-Based Thermal Imaging in the Management and Monitoring of Microvascular Flap Procedures: A Systematic Review and Meta-Analysis. Ann. Plast. Surg. 90 (6S Suppl 4), S420–S425. 10.1097/SAP.0000000000003343 (2023). Kwasnicki, R. M., Noakes, A. J., Banhidy, N. & Hettiaratchy, S. Quantifying the Limitations of Clinical and Technology-based Flap Monitoring Strategies using a Systematic Thematic Analysis. Plast. Reconstr. Surg. - Global Open. 9 (7), e3663. 10.1097/GOX.0000000000003663 (2021). Kyriacou, P. A. Photoplethysmography: Technology, Signal Analysis and Applications (Elsevier, AP Academic, 2022). Charlton, P. H. et al. The 2023 wearable photoplethysmography roadmap. Physiol. Meas. 44 (11), 111001. 10.1088/1361-6579/acead2 (2023). Zaman, T., Kyriacou, P. A. & Pal, S. K. Development of a reflectance photoplethysmographic sensor used for the assessment of free flap perfusion. In: 2011 Annual International Conference of the IEEE Engineering in Medicine and Biology Society . IEEE; :4006–4009. (2011). 10.1109/IEMBS.2011.6090995 Kyriacou, P. A., Zaman, T. & Pal, S. K. Photoplethysmography in postoperative monitoring of deep inferior epigastric perforator (DIEP) free flaps. Physiol. Meas. 41 (12), 124001. 10.1088/1361-6579/abc4c8 (2020). Hong, K. S., Park, K. T. & Ahn, J. M. Aging Index using Photoplethysmography for a Healthcare Device: Comparison with Brachial-Ankle Pulse Wave Velocity. Healthc. Inf. Res. 21 (1), 30. 10.4258/hir.2015.21.1.30 (2015). Grabovskis, A., Marcinkevics, Z., Rubins, U. & Kviesis-Kipge, E. Effect of probe contact pressure on the photoplethysmographic assessment of conduit artery stiffness. J. Biomed. Opt. 18 (2), 027004. 10.1117/1.JBO.18.2.027004 (2013). Millasseau, S. C., Ritter, J. M., Takazawa, K. & Chowienczyk, P. J. Contour analysis of the photoplethysmographic pulse measured at the finger. J. Hypertens. 24 (8), 1449–1456. 10.1097/01.hjh.0000239277.05068.87 (2006). Budidha, K., Rybynok, V. & Kyriacou, P. A. Design and Development of a Modular, Multichannel Photoplethysmography System. IEEE Trans. Instrum. Meas. 67 (8), 1954–1965. 10.1109/TIM.2018.2810643 (2018). Ferizoli, R., Karimpour, P., May, J. M. & Kyriacou, P. A. Arterial stiffness assessment using PPG feature extraction and significance testing in an in vitro cardiovascular system. Sci. Rep. 14 (1), 2024. 10.1038/s41598-024-51395-y (2024). Karimpour, P., Ferizoli, R., May, J. M. & Kyriacou, P. A. Customisable Silicone Vessels and Tissue Phantoms for In Vitro Photoplethysmography Investigations into Cardiovascular Disease. Sensors 24 (5), 1681. 10.3390/s24051681 (2024). Cappelletti, S. et al. Non-invasive estimation of vascular compliance and distensibility in the arm vessels: a novel ultrasound-based protocol. Quant. Imaging Med. Surg. 12 (7), 3515–3527. 10.21037/qims-21-987 (2022). Nomoni, M., May, J. M. & Kyriacou, P. A. Novel Polydimethylsiloxane (PDMS) Pulsatile Vascular Tissue Phantoms for the In-Vitro Investigation of Light Tissue Interaction in Photoplethysmography. Sensors 20 (15), 4246. 10.3390/s20154246 (2020). Chan, G. et al. Multi-Site Photoplethysmography Technology for Blood Pressure Assessment: Challenges and Recommendations. JCM 8 (11), 1827. 10.3390/jcm8111827 (2019). Ferizoli, R., Karimpour, P., May, J. M. & Kyriacou, P. A. A Bilateral In Vitro Model for Cardiovascular Disease Investigations Using Photoplethysmography Sensors. In: 2023 IEEE BioSensors Conference (BioSensors) . IEEE; :1–4. (2023). 10.1109/BioSensors58001.2023.10281067 UV-Vis Spectrum of Methylene Blue Solution, with Maximu|Open-i. Available online: https://openi.nlm.nih.gov/detailedresult?img=PMC2808005_ijms-10-04707f6&req=4 (accessed on 16 January 2025). Matsumine, H. et al. Improvement of the salvage-rate of flap after venous thrombosis with intraparenchymatous venous pressure monitoring. Microsurgery 38 (5), 498–503. 10.1002/micr.30293 (2018). Saiga, A., Kubota, Y., Yamaji, Y. & Mitsukawa, N. Intraflap Vascular Catheterization Method for Monitoring, Prevention, and Intervention of Thrombogenesis in Free-Flap Surgery. Ann. Plast. Surg. 88 (1), 68–73. 10.1097/SAP.0000000000003049 (2022). Goda, M. Á., Charlton, P. H. & Behar, J. A. pyPPG: a Python toolbox for comprehensive photoplethysmography signal analysis. Physiol. Meas. 45 (4), 045001. 10.1088/1361-6579/ad33a2 (2024). Liang, Y., Chen, Z., Ward, R. & Elgendi, M. Photoplethysmography and Deep Learning: Enhancing Hypertension Risk Stratification. Biosensors 8 (4), 101. 10.3390/bios8040101 (2018). Oppenheim, A. V. & Schafer, R. W. Discrete-Time Signal Processing (Pearson, 2010). https://books.google.co.uk/books?id=mYsoAQAAMAAJ Leys, C., Ley, C., Klein, O., Bernard, P. & Licata, L. Detecting outliers: Do not use standard deviation around the mean, use absolute deviation around the median. J. Exp. Soc. Psychol. 49 (4), 764–766. 10.1016/j.jesp.2013.03.013 (2013). Miller, J. Short Report: Reaction Time Analysis with Outlier Exclusion: Bias Varies with Sample Size. Q. J. Experimental Psychol. Sect. A . 43 (4), 907–912. 10.1080/14640749108400962 (1991). Ram, M. R., Madhav, K. V., Krishna, E. H., Komalla, N. R. & Reddy, K. A. A Novel Approach for Motion Artifact Reduction in PPG Signals Based on AS-LMS Adaptive Filter. IEEE Trans. Instrum. Meas. 61 (5), 1445–1457. 10.1109/TIM.2011.2175832 (2012). Wang, L., Lo, B. P. & Yang, G. Z. Multichannel Reflective PPG Earpiece Sensor With Passive Motion Cancellation. IEEE Trans. Biomed. Circuits Syst. 1 (4), 235–241. 10.1109/TBCAS.2007.910900 (2007). Jang, D. G., Park, S., Hahn, M. & Park, S. H. A Real-Time Pulse Peak Detection Algorithm for the Photoplethysmogram. IJEEE Published online 2014 :45–49. 10.12720/ijeee.2.1.45-49 Millasseau, S. C., Ritter, J. M., Takazawa, K. & Chowienczyk, P. J. Contour analysis of the photoplethysmographic pulse measured at the finger. J. Hypertens. 24 (8), 1449–1456. 10.1097/01.hjh.0000239277.05068.87 (2006). Pang, J. et al. Nonparametric bootstrap methods for interval estimation of the area under the ROC curve with correlated diagnostic test data: application to whole-virus ELISA testing in swine. Front. Vet. Sci. 10 , 1274786. 10.3389/fvets.2023.1274786 (2023). Falanga, V. & Bucalo, B. Use of a durometer to assess skin hardness. J. Am. Acad. Dermatol. 29 (1), 47–51. 10.1016/0190-9622(93)70150-R (1993). Pereira, T. et al. A Supervised Approach to Robust Photoplethysmography Quality Assessment. IEEE J. Biomed. Health Inf. 24 (3), 649–657. 10.1109/JBHI.2019.2909065 (2020). Charlton, P. H., Marozas, V., Mejía-Mejía, E., Kyriacou, P. A. & Mant, J. Determinants of photoplethysmography signal quality at the wrist. PLOS Digit. Health . 4 (6), e0000585. 10.1371/journal.pdig.0000585 (2025). Reconstructive, B. A., Surgeons, P. & Surgery, A. of A of B. Oncoplastic Breast Reconstruction Guidelines for Best Practice. Published online 2012. http://www.bapras.org.uk/docs/default-source/commissioning-and-policy/final-oncoplastic-guidelines–healthcare-professionals.pdf?sfvrsn=0 Creech, B. & Miller, S. Evaluation of circulation in skin flaps. Skin Flaps Boston: Little Brown ; 21 . (1975). Swartz, W. M., Jones, N. F., Cherup, L., Klein, A. & Pittsburgh Direct Monitoring of Microvascular Anastomoses with the 20-MHz Ultrasonic Doppler Probe: An Experimental and Clinical Study. Plast. Reconstr. Surg. 81 (2), 149–158. 10.1097/00006534-198802000-00001 (1988). Geoffrey, G. & Hallock Chapter 3: Flap selection. In: Flaps and Reconstructive Surgery. Elsevier Inc; :17. (2009). Tomioka, Y. et al. Wearable, wireless, multi-sensor device for monitoring tissue circulation after free-tissue transplantation: a multicentre clinical trial. Sci. Rep. 12 (1), 16532. 10.1038/s41598-022-21007-8 (2022). Min, K., Hong, J. P. & Suh, H. P. Risk Factors for Partial Flap Loss in a Free Flap: A 12-Year Retrospective Study of Anterolateral Thigh Free Flaps in 303 Lower Extremity Cases. Plast. Reconstr. Surg. 150 (5), 1071e–1081e. 10.1097/PRS.0000000000009646 (2022). Macrae, R. A., Miller, K. & Doyle, B. J. Methods in Mechanical Testing of Arterial Tissue: A Review. Strain 52 (5), 380–399. 10.1111/str.12183 (2016). Pajic-Lijakovic, I., Milivojevic, M. & McClintock, P. V. E. Compressibility of biological systems: the viscoelastic Poisson’s ratio. Adv. Physics: X . 10 (1), 2440023. 10.1080/23746149.2024.2440023 (2025). Ayhan, S. et al. Correlation between vessel diameters of superficial and deep inferior epigastric systems: Doppler ultrasound assessment. J. Plast. Reconstr. Aesthetic Surg. 62 (9), 1140–1147. 10.1016/j.bjps.2008.02.012 (2009). Falanga, V. & Bucalo, B. Use of a durometer to assess skin hardness. J. Am. Acad. Dermatol. 29 (1), 47–51. 10.1016/0190-9622(93)70150-R (1993). Oprişan, C. et al. Experimental determination of the Young’s modulus for the fingers with application in prehension systems for small cylindrical objects. IOP Conf. Ser: Mater. Sci. Eng. 147 , 012058. 10.1088/1757-899X/147/1/012058 (2016). Watanabe, Y., Yamane, N., Sathyanarayana, A., Mishra, V. & Goodwin, M. S. Beyond Motion Artifacts: Optimizing PPG Preprocessing for Accurate Pulse Rate Variability Estimation. arXiv . Preprint posted online 2025. 10.48550/ARXIV.2510.06158 Sahan, B., Yolcu, C. & Sahin, S. Energy consumption and boot time performances of Kentkart industrial tablet on Android platform. In: International Conference on Engineering and Technology (ICET) . IEEE; 2017:1–4. IEEE; 2017:1–4. (2017). 10.1109/ICEngTechnol.2017.8308177 Joseph, G., Joseph, A., Titus, G., Thomas, R. M. & Jose, D. Photoplethysmogram (PPG) signal analysis and wavelet de-noising. In: 2014 Annual International Conference on Emerging Research Areas: Magnetics, Machines and Drives (AICERA/iCMMD) . IEEE; :1–5. (2014). 10.1109/AICERA.2014.6908199 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8080454","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":550149654,"identity":"434f7a24-2c29-49a7-991c-3a389b5a1427","order_by":0,"name":"Hiroki Kodama","email":"data:image/png;base64,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","orcid":"","institution":"St George's, University of London","correspondingAuthor":true,"prefix":"","firstName":"Hiroki","middleName":"","lastName":"Kodama","suffix":""},{"id":550149655,"identity":"dce4bb1a-3f76-4dfd-9913-19fe35c6462a","order_by":1,"name":"James May","email":"","orcid":"","institution":"St George's, University of London","correspondingAuthor":false,"prefix":"","firstName":"James","middleName":"","lastName":"May","suffix":""},{"id":550149656,"identity":"bec84ee1-3539-4fd3-bbaa-595ba2dc185d","order_by":2,"name":"Dariush Nikkhah","email":"","orcid":"","institution":"Royal Free London NHS Foundation Trust","correspondingAuthor":false,"prefix":"","firstName":"Dariush","middleName":"","lastName":"Nikkhah","suffix":""},{"id":550149657,"identity":"e5b94f3f-d81e-4bc4-b346-f72e4a7b26f8","order_by":3,"name":"Panicos A Kyriacou","email":"","orcid":"","institution":"St George's, University of London","correspondingAuthor":false,"prefix":"","firstName":"Panicos","middleName":"A","lastName":"Kyriacou","suffix":""}],"badges":[],"createdAt":"2025-11-10 20:38:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8080454/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8080454/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":96918489,"identity":"ad16e9e2-e535-41d1-8529-f016e4e69939","added_by":"auto","created_at":"2025-11-27 14:12:01","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2406928,"visible":true,"origin":"","legend":"","description":"","filename":"FlapphantomscriptSciRepfinal.docx","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/2df0030ffbd8df66dbbb3986.docx"},{"id":96845479,"identity":"d11ef9b4-31ce-4233-b6c9-2f5c15d30281","added_by":"auto","created_at":"2025-11-26 16:33:46","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6546,"visible":true,"origin":"","legend":"","description":"","filename":"010706306bf445f4a6a710b4647d24bc.json","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/b3e0025bb330e4f81589b0bc.json"},{"id":96919608,"identity":"2e94715e-b749-492e-bfeb-3e8562d7947e","added_by":"auto","created_at":"2025-11-27 14:14:12","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":155737,"visible":true,"origin":"","legend":"","description":"","filename":"010706306bf445f4a6a710b4647d24bc1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/42d4209b59e0ebf07824345d.xml"},{"id":96919139,"identity":"a7ba5cc1-2366-434d-a9c4-86cab30f35cf","added_by":"auto","created_at":"2025-11-27 14:13:14","extension":"emf","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1542364,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.emf","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/0875f0201721091eba396af6.emf"},{"id":96845492,"identity":"3d37acd5-2f8e-4688-a80e-99f4816b8007","added_by":"auto","created_at":"2025-11-26 16:33:46","extension":"emf","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1348308,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.emf","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/483153be3b89adf42eb4257f.emf"},{"id":96845490,"identity":"f3ba8ba3-197b-48be-a3b2-c5028043a2c7","added_by":"auto","created_at":"2025-11-26 16:33:46","extension":"emf","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":930868,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.emf","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/7af889ac97904a89b8ce82ae.emf"},{"id":96845494,"identity":"bcb34dc1-1365-43e3-8133-a30ce26acfd5","added_by":"auto","created_at":"2025-11-26 16:33:46","extension":"emf","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2062372,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.emf","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/b3126ebcc508ee83ba15b023.emf"},{"id":96845493,"identity":"1dc63f6b-d963-458a-851e-4564359fcd25","added_by":"auto","created_at":"2025-11-26 16:33:46","extension":"emf","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1959060,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.emf","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/2acc8789097b6c3726fb2fdb.emf"},{"id":96845495,"identity":"334bdc93-6ff3-4fc8-ad16-64e0fa8a97ad","added_by":"auto","created_at":"2025-11-26 16:33:46","extension":"emf","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1711840,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.emf","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/365e83de4f9036846a1fc6c4.emf"},{"id":96845487,"identity":"73a82c25-f56e-4cef-bfef-a9e807b2684e","added_by":"auto","created_at":"2025-11-26 16:33:46","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":18268,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/74d9e786be4d78c1300cbf39.png"},{"id":96845489,"identity":"bb9a9a2b-1fac-4270-871e-375caf3b770f","added_by":"auto","created_at":"2025-11-26 16:33:46","extension":"xml","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":155702,"visible":true,"origin":"","legend":"","description":"","filename":"010706306bf445f4a6a710b4647d24bc1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/31bf3f3939a3223638713a9d.xml"},{"id":96919728,"identity":"bb92de28-ae45-4b62-a99b-22b8321ffce8","added_by":"auto","created_at":"2025-11-27 14:14:22","extension":"html","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":168390,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/38612166555288eb541a650a.html"},{"id":96845478,"identity":"94c93ca8-5838-466b-9821-b4a400348d27","added_by":"auto","created_at":"2025-11-26 16:33:46","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":44020,"visible":true,"origin":"","legend":"\u003cp\u003e(a) Details of the Fabricated Phantom and Experimental Conditions. (b)Schematic Diagram of the Cardiovascular System. A PPG sensor is applied to a tissue phantom containing an artery connected to a pump and a vein connected to a closed circuit.\u003c/p\u003e","description":"","filename":"Picture1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/801e85d4c3311d4ebb1a822a.jpg"},{"id":96845480,"identity":"31397751-5a8f-466a-a989-052b79384398","added_by":"auto","created_at":"2025-11-26 16:33:46","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":25395,"visible":true,"origin":"","legend":"\u003cp\u003eStatistical analysis workflow for PPG signal evaluation and classification\u003c/p\u003e","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/c9aa00e9f7bdd54b790c6f63.png"},{"id":96845482,"identity":"29a7c8e1-95d3-4ad9-908f-15bc81fed9ac","added_by":"auto","created_at":"2025-11-26 16:33:46","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":15884,"visible":true,"origin":"","legend":"\u003cp\u003eExample of PPG morphological features\u003c/p\u003e","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/f21c6f3c40920333c3801726.png"},{"id":96845483,"identity":"a5301cd6-d555-4374-9fad-5d4f8c7c9763","added_by":"auto","created_at":"2025-11-26 16:33:46","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":29173,"visible":true,"origin":"","legend":"\u003cp\u003e(a) Calculation of the mechanical properties of the created vascular phantom. d: Wall thickness, OD: Outer Diameter, ID: Inner Diameter, CSA: Cross-sectional Area, C: Compliance. (b) At Hardener (H) = 0.6, the ΔCSA was 6.62–8.09 mm\u003csup\u003e2\u003c/sup\u003e (+22.5%) at 40–80 mmHg and 7.34–9.14 mm\u003csup\u003e2\u003c/sup\u003e (+24.4%) at 60–120 mmHg. (c) At H = 0.6, C was 3.71×10\u003csup\u003e-4\u003c/sup\u003e cm\u003csup\u003e2\u003c/sup\u003e/mmHg at 40-80 mmHg and 2.99×10\u003csup\u003e-4\u003c/sup\u003e cm\u003csup\u003e2\u003c/sup\u003e/mmHg at 60-120 mmHg. (d) At Deadener (D) = 0.4, ΔCSA was 9.36–12.5 mm\u003csup\u003e2\u003c/sup\u003e (+33.6%). (e) At D = 0.4, C was 15.7×10-4 cm\u003csup\u003e2\u003c/sup\u003e/mmHg at 10-30 mmHg.\u003c/p\u003e","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/233bc9486b2c1623342ffc13.png"},{"id":96845485,"identity":"fd027b29-6783-4bc4-bb83-6fa0f72fe00b","added_by":"auto","created_at":"2025-11-26 16:33:46","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":44754,"visible":true,"origin":"","legend":"\u003cp\u003e(a) The intraluminal pressure was controlled as intended across all graphs. (b) Regular waveforms were obtained, except for the Red channel at deep positions. However, the Perfusion Index (PI: AC/DC ratio) was within the acceptable analysis range, measuring between 0.12% and 0.43%.\u003c/p\u003e","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/3ae74e4dd0d303d6b350e27d.png"},{"id":96845484,"identity":"4b9b847e-8818-41a9-964c-74e256b75394","added_by":"auto","created_at":"2025-11-26 16:33:46","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":21344,"visible":true,"origin":"","legend":"\u003cp\u003eAll conditions, with the exception of Congestion at 21 mm, were shown to be detectable with high accuracy. Specifically, the differentiation between Normal and Ischaemia/Congestion was highly accurate using IR Amplitude, and the distinction between Ischaemia and Congestion was accurately achieved using IR Upslope Length.\u003c/p\u003e","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/2fc023925a75aa2caa13317d.png"},{"id":103205359,"identity":"29dc5f74-a3b9-4df3-8aae-87c33e446fd3","added_by":"auto","created_at":"2026-02-23 07:12:10","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1164828,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8080454/v1/06d9e96e-98b3-4b15-be98-36eff3e8c33a.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Diagnostic Performance of Photoplethysmography (PPG) for Early Vascular Compromise in a Customisable In Vitro Flap Model","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eFree flaps are tissue units transferred from a donor to a recipient site while maintaining their original blood supply via microvascular anastomosis (connection between two blood vessels). Free flap reconstruction represents a cornerstone method for managing large and complex defects, offering superior functional and aesthetic outcomes in diverse anatomical locations, including the head and neck\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e, breast\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, and extremities\u003csup\u003e\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. The procedure's growing importance is evidenced by national data; for example, a 2020 report from the United States indicated that the number of free flap procedures performed for breast reconstruction reached 23,324, representing a 75% increase over the preceding two decades\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eDespite the substantial benefits of microvascular tissue transplantation, blood flow deficits remain a significant concern, occurring in approximately 3% to 5% of cases\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Vascular thrombosis is the principal risk factor, typically manifesting within the first 48 hours post-surgery\u003csup\u003e\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e. A severe consequence of delayed diagnosis is a significantly reduced probability of free flap salvage. Furthermore, the need for a second flap to address failure is reported to incur an additional cost of \u003cspan\u003e$\u003c/span\u003e174,000\u003csup\u003e9\u003c/sup\u003e. Crucially, historical studies report that prompt salvage surgeries initiated within the first 24 hours of vascular compromise successfully save 70% to 80% of affected flaps\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e,\u003cspan additionalcitationids=\"CR11\" citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eThe traditional method of clinical assessment remains the established standard for evaluating free flaps and achieving early detection of compromise. This evaluation includes measurements of skin colour, capillary refill time, tissue fullness (turgour) caused by fluid build up, perceptible by the physical examination of how hard or soft the tissue is, temperature, and pinprick response. However, these observations are labour-intensive, as they must be performed manually and frequently\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. Furthermore, these characteristics are qualitatively and subjectively monitored, largely depending on the expertise of the individual physician or nurse. Because different personnel often monitor the flap, inconsistencies may arise in these observations, risking the oversight of subtle changes\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eVarious monitoring devices are currently available on the market. Pinpoint measurement, devices such as Near-Infrared Spectroscopy (NIRS)\u003csup\u003e\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e, Laser Doppler Flowmetry (LDF)\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e, implantable Doppler probes\u003csup\u003e\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e, flow couplers\u003csup\u003e\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u003c/sup\u003e, and microdialysis\u003csup\u003e\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e are utilised. For image-based assessment, techniques such as Indocyanine Green Fluorescence Angiography\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e and thermography\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e are available. Each of these technologies possesses inherent strengths and limitations, and a universal consensus regarding their optimal clinical use does not currently exist\u003csup\u003e\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003ePhotoplethysmography (PPG), alongside NIRS, is an optical sensing technology that allows for non-invasive and continuous monitoring. While NIRS measures tissue-level oxygenation, PPG monitors changes in vascular volume. A PPG sensor primarily consists of two components: a light source, typically a Light Emitting Diode (LED), and a photodetector (such as a photodiode or phototransistor) that measures the amount of received light\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. Depending on the body site where the sensor is placed, it operates in either transmission or reflectance mode. In transmission mode, red and infrared light from the LED passes through a body part and is detected by a photodetector positioned on the opposite side. In reflectance mode, the photodetector measures the amount of light reflected back from the tissue\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eDue to its small size and low cost, PPG has long been used as the fundamental technology for pulse oximetry in clinical settings. Recent technological advancements have led to this technique being widely incorporated into wearable devices such as smartwatches and rings, and its clinical significance is rapidly increasing\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. This sensor technology enables continuous data acquisition, making it possible to detect abnormalities in a timely manner that cannot be achieved with static images. However, direct evidence regarding PPG's specific diagnostic capability in free flap monitoring remains limited\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eAn ideal PPG signal is characterised by a systolic peak and a diastolic peak, separated by the dicrotic notch. However, in practice, the diastolic peak is frequently absent, particularly in elderly patients\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e,\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. In this study, we adopted a feature extraction algorithm based solely on the systolic peak from the PPG waveform, and the features were used to identify the correlation with early ischemia and congestion. Furthermore, the features analysed rely on the original filtered PPG waveform, rather than its derivative. Consequently, this correlation is applicable even when the second derivative reference points used in existing pulse wave analysis methods are unavailable\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eThis paper aims to investigate the diagnostic capability of PPG for early blood flow changes in free flaps by evaluating PPG signals acquired from an \u003cem\u003ein vitro\u003c/em\u003e vascular system. By demonstrating the utility of various pulse wave characteristics, this work is expected to inform future monitoring technologies. The research is predicated on the hypothesis that specific morphological features of the pulse wave serve as discernible indicators of early blood flow changes.\u003c/p\u003e"},{"header":"2. Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.1. Phantom Creation and Experimental Setup\u003c/h2\u003e\u003cp\u003eThe experimental system was built using established Zen PPG hardware\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. To simulate free flap vasculature, custom-made silicone vessels were fabricated according to previous methods\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e, accurately mimicking the ΔCSA (Change in Cross-Sectional Area) and compliance reported in the literature\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. The general experimental circuit for PPG validation was modified to specifically model flap circulation, incorporating the fabricated vessels at varying depths within a tissue phantom. Intravascular pressure was precisely controlled to simulate the small artery and vein under specific conditions, including normal flow, early ischemia, and early congestion. This setup allowed for an \u003cem\u003ein vitro\u003c/em\u003e assessment by creating precise and reproducible flow abnormalities.\u003c/p\u003e\u003cdiv id=\"Sec4\" class=\"Section3\"\u003e\u003ch2\u003e2.1.1. Phantom Creation\u003c/h2\u003e\u003cp\u003eThe phantom creation was based on previous research conducted in our laboratory\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e,\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. The mechanical properties of the vessels were reproduced by mixing equal parts of Gel-00 (Polytek Development Corp., Easton, PA, USA) Part A and Part B with appropriate ratios of deadener (D) and hardener (H), along with a 2% retarder\u003csup\u003e31\u003c/sup\u003e. The wall thickness (d) and outer diameter (OD) of the fabricated vessels were measured. Based on the degree of expansion in response to the appropriate internal pressure, ΔCSA and compliance were calculated to determine the optimal blend ratio for mimicking human tissue.\u003c/p\u003e\u003cp\u003eFour separate vascular tissue phantoms were fabricated by embedding the vessels (artery and vein) within a surrounding silicone matrix. This process, which followed the previously established casting procedure\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e, positioned the vessels at depths of 3, 9, 15, and 21 mm from the phantom's surface (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e (a)).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section3\"\u003e\u003ch2\u003e2.1.2. \u003cem\u003eIn Vitro\u003c/em\u003e Cardiovascular System Configuration\u003c/h2\u003e\u003cp\u003eThe fabricated tissue phantom was incorporated into an \u003cem\u003ein vitro\u003c/em\u003e vascular system, replicating the systemic circulation system of the human body. The phantom was positioned within a dedicated sensor casing to ensure consistent monitoring. A reflectance PPG sensor, featuring Red (660 nm) and Infrared (940 nm) LEDs, was placed on the underside of the phantom. This sensor was connected to a custom dual-channel PPG acquisition system, developed by the Research Centre for Biomedical Engineering (RCBE) group at City St George\u0026rsquo;s, University of London\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e,\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e, which operates at a sampling rate of 2000 Hz. The signals were displayed and recorded using LabVIEW (version 2023 Q1, National Instruments, Austin, Texas).\u003c/p\u003e\u003cp\u003eTo simulate blood flow, a liquid composed of deionized water and methylene blue powder (Thermo Fisher Scientific, UK)\u003csup\u003e\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u003c/sup\u003e was circulated throughout the system using a pulsating pump (PD-1100, BDC Laboratories, Wheat Ridge, CO) operating at 60\u0026ndash;65 bpm. The phantom vein was assumed to have minimal pulsation and was therefore connected to a closed circuit with adjustable internal pressure (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e (b)).\u003c/p\u003e\u003cp\u003eBased on previous studies, arterial internal pressure was set to 40\u0026ndash;80 mmHg for the normal and congested state and 25\u0026ndash;50 mmHg for the ischemic state\u003csup\u003e\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e,\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. The adjustment of internal pressure was done by adjusting a clamp within the circuit. The venous internal pressure, on the other hand, was set to 10 mmHg for the normal and ischemic states and 30 mmHg for the congested state, which was adjusted by injecting fluid via a syringe through a valve (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e (b)).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003e2.2. Signal Processing and Statistical Analysis\u003c/h2\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e illustrates the flow of procedures from signal processing through statistical analysis. The acquired signals were assessed for quality by two independent experts, using the following criteria: 'Excellent' (clear systolic and diastolic waves), 'Acceptable' (heart rate detectable), and 'Poor' (heart rate is difficult to detect). The signals were further evaluated using the Perfusion Index (AC/DC Ratio) before proceeding to the analysis stage.\u003c/p\u003e\u003cp\u003ePPG signal preprocessing was performed following a standard signal processing pipeline consisting of four sequential stages: (1) bandpass filtering, (2) artifact detection, (3) artifact removal, and (4) baseline correction.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eTo preserve the primary physiological components of PPG signals (cardiac-related frequency components) while removing high-frequency noise and low-frequency drift, a bandpass filter with cutoff frequencies of 0.5\u0026ndash;10 Hz was applied\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e. A 4th-order Chebyshev Type II filter was employed for filter design\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. The stopband attenuation was set to 40 dB, and zero-phase filtering (forward-backward filtering) was implemented to avoid phase distortion\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eFollowing filtering, artifacts were detected using the MAD (Median Absolute Deviation) method, a statistical outlier detection technique\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. The modified Z-score was calculated as follows\u0026hellip;\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:Modified\\:Z-score=0.6745\\times\\:\\frac{(Xi-median)}{MAD}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u0026hellip;where MAD\u0026thinsp;=\u0026thinsp;median(|Xi - median|), and 0.6745 is the conversion factor from MAD to standard deviation under the assumption of normal distribution. Data points with an absolute modified Z-score exceeding 3.5 were identified as artifacts\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eDetected artifact points were replaced using linear interpolation\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. Specifically, values at artifact positions were estimated by linear interpolation using valid signal points that were not identified as artifacts.\u003c/p\u003e\u003cp\u003eIn this study, baseline estimation and correction were performed using the cubic spline interpolation method\u003csup\u003e\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e,\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e\u003c/sup\u003e. First, the baseline component was estimated as the lower envelope of the signal using a moving minimum filter with a window size corresponding to approximately one cardiac cycle (0.8 s, 1600 sample points). Next, a cubic spline function was fitted to the estimated baseline points with a smoothing parameter of s\u0026thinsp;=\u0026thinsp;0.01 \u0026times; N (where N is the number of sample points) to obtain a smoothed baseline curve. Finally, the baseline-corrected PPG signal was obtained by subtracting this baseline curve from the original signal.\u003c/p\u003e\u003cp\u003eThrough these preprocessing steps, high-quality signals with reduced noise and baseline variations were obtained for both Red and IR channel PPG signals.\u003c/p\u003e\u003cp\u003ePPG features were extracted from the pulse waves (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e)\u003csup\u003e\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eTable of PPG features\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"2\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFeature (abbreviation)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDescription\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eArea Under Curve (AUC)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eArea of pulse\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSystolic Area Under Curve (S-AUC)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eArea of the systolic phase of the pulse\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDiastolic Area Under Curve (D-AUC)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eArea of the diastolic phase of the pulse\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eArea Under Curve Ratio (AUC\u0026nbsp;Ratio)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eRatio between systolic and diastolic areas\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRise Time\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eTime taken for signal to rise from trough to peak\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDecay Time\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eTime taken for signal to fall from peak to trough\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRise-Decay Time Ratio\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eRatio of decay time to rise time\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAmplitude\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMedian amplitude\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUpslope Length\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eLength of tangent from trough to peak\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDownslope Length\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eLength of tangent from peak to trough\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUpslope\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eGradient of pulse on rising edge\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDownslope\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eGradient of pulse on falling edge\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOnset-End Slope\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eRate of change of the pulse over the entire pulse length\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSlope Ratio\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eRising slope gradient (upslope) compared to falling slope gradient (downslope)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLength-Height Ratio\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eRatio of the height of a pulse compared to its width\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSlope Length Ratio\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eRatio of the upslope length of a pulse compared to its downslope length\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUpslope Length Ratio\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eProportion of the total pulse length that is made up of the rising phase\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDownslope Length Ratio\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eProportion of the total pulse length that is made up of the falling phase\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eStart Datum Area\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eArea between PPG signal and the tangent from trough to peak (systolic phase)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eEnd Datum Area\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eArea between PPG signal and the tangent from peak to trough (diastolic phase)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDatum Area Ratio\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eRatio between systolic and diastolic datum areas\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMax Start Datum Difference\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMaximum length between PPG signal and peak-trough tangent (systolic phase)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMax End Datum Difference\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMaximum length between PPG signal and trough-peak tangent (diastolic phase)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMedian Start Datum Difference\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMedian length between PPG signal and peak-trough tangent (systolic phase)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMedian End Datum Difference\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eMedian length between PPG signal and trough-peak tangent (diastolic phase)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePulse Width\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWidth of the entire pulse of one cycle (measured at 50% of height)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSystolic Width\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWidth of the systolic phase of the pulse (measured at 50% of height)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDiastolic Width\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWidth of the diastolic phase of the pulse (measured at 50% of height)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eWidth Ratio\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eRatio between the systolic and diastolic widths (measured at 50% of height)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVariance\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDeviation of signal from its mean value\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSkewness\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDegree of symmetry of a pulse\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKurtosis\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDegree of sharpness of a pulse\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSignal-to-Noise Ratio (SNR)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eLevel of PPG signal compared to noise\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eZero-Crossing Rate (ZCR)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNumber of times per second that the PPG signal crosses zero\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePeak-To-Instantaneous Ratio (PIR)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDegree of change in amplitude over time\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThe diagnostic performance for differentiating the Normal, early Ischemic, and early Congested states was evaluated using a comprehensive feature set. Receiver Operating Characteristic (ROC) analysis was performed for each pairwise comparison, wavelength, and depth to complement group-level p-values with a measure of classification ability. Diagnostic accuracy was quantified by generating these ROC curves based on sensitivity and false-positive rates, with the Area Under the Curve (AUC) serving as the final performance metric to quantify how well the normalised amplitude could distinguish between two states. To ensure the robustness of the AUC estimates, 95% confidence intervals (CIs) were derived using a bootstrap resampling method (N\u0026thinsp;=\u0026thinsp;200 resamplings). Furthermore, the optimal threshold (cutoff) for each feature and each pair was identified by maximising Youden's J statistic (sensitivity\u0026thinsp;+\u0026thinsp;specificity \u0026minus;\u0026thinsp;1), and the corresponding sensitivity and specificity were recorded. Features were defined as possessing diagnostic performance if their Area Under the Curve (AUC) exceeded 0.7\u003csup\u003e48\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eAll statistical analyses and visualizations were performed in Python 3.13.0.\u003c/p\u003e\u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e3.1. Phantom Fabrication and Mechanical Characterization\u003c/h2\u003e\u003cp\u003eThe change inΔCSA and compliance were calculated as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e(a). The blend ratio for the arterial phantom was adjusted to approximate a ΔCSA range of 23.5%\u0026ndash;25.9% and a compliance range of 1.3\u0026ndash;3.52\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;4\u003c/sup\u003e cm\u003csup\u003e2\u003c/sup\u003e /mmHg across pressure changes of 40\u0026ndash;80 mmHg and 60\u0026ndash;120 mmHg\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. The final ratio of the Hardener was determined to be 0.6, resulting in an initial internal diameter (ID\u003csub\u003e0\u003c/sub\u003e) of 2.72 mm(Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e(b)). Similarly, the blend ratio for the venous phantom was adjusted to approximate a ΔCSA range of 27.2%\u0026ndash;41.8% and a compliance range of 5.6\u0026ndash;17.4\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;4\u003c/sup\u003ecm\u003csup\u003e2\u003c/sup\u003e /mmHg across pressure changes of 10\u0026ndash;30 mmHg\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. The final ratio of the Deadener (D) was determined to be 0.4, resulting in an initial ID\u003csub\u003e0\u003c/sub\u003e of 3.19 mm (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e(c)). These artery and vein phantoms were then embedded 3, 9, 15, 21 mm deep within a subcutaneous tissue phantom (ratio of D was 1) created based on a previous report\u003csup\u003e\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e,\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e,\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e3.2. Signal Acquisition\u003c/h2\u003e\u003cp\u003eAcross each wavelength, depth, and condition, the appropriate internal pressures were successfully replicated, and stable waveforms were acquired (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e3.3. Signal Quality Assessment\u003c/h2\u003e\u003cp\u003eSignal quality, assessed independently by two PhD students, was deemed 'Excellent' (clear systolic and diastolic waves) or 'Acceptable' (heart rate detectable) for the Red signal up to a depth of 9 mm and for all IR signals\u003csup\u003e\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e\u003c/sup\u003e. Furthermore, the calculated Perfusion Index (PI: AC/DC Ratio) was 0.13%\u0026ndash;0.43% (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), which falls within the previously reported range of 0.12%\u0026ndash;0.76%\u003csup\u003e51\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003ePerfusion Index (PI) at each depth\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDepth\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eState\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eChannel\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eQuality\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003ePI (%)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e\u003cp\u003e3\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eNormal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eExcellent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.35\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eIR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eExcellent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.43\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eIschaemia\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eExcellent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.22\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eIR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eExcellent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.39\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eCongestion\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eExcellent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.43\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eIR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eExcellent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.38\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e\u003cp\u003e9\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eNormal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAcceptable\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.17\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eIR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eExcellent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.21\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eIschaemia\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAcceptable\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.14\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eIR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eExcellent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.16\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eCongestion\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAcceptable\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.18\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eIR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eExcellent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.22\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e\u003cp\u003e15\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eNormal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eUnfit\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.16\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eIR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eExcellent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.16\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eIschaemia\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eUnfit\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.15\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eIR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eExcellent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.13\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eCongestion\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eUnfit\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.16\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eIR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eExcellent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.17\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"5\" rowspan=\"6\"\u003e\u003cp\u003e21\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eNormal\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eUnfit\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.18\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eIR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eExcellent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.16\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eIschaemia\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eUnfit\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.17\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eIR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAcceptable\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.18\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e\u003cp\u003eCongestion\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eRed\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eUnfit\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.19\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eIR\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eExcellent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.14\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e3.4. Diagnostic Ability\u003c/h2\u003e\u003cp\u003eDiagnostic performance was defined by an Area Under the Curve (AUC) exceeding 0.7. These diagnostically useful features are presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e, segmented by pairwise condition and depth. Amplitude proved to be the most reliable feature for distinguishing between normal and abnormal perfusion states. While detecting early congestion at the deepest tissue levels presented a challenge, various other extracted features collectively enabled discrimination for all remaining diagnostic pairs.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eThe required frequency for free flap assessment, with frequency as high as every hour to every 30 minutes\u003csup\u003e\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e\u003c/sup\u003e, has historically placed a substantial burden on healthcare systems, primarily relying on in-house resident physicians\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e. In the current environment, which prioritises strict limitations on working hours, this manual monitoring approach is reaching its practical limit\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e, creating a strong desire for reliable monitoring devices. Despite this, a culture of reliance on manual clinical checks persists.\u003c/p\u003e\u003cp\u003eCreech and Miller\u003csup\u003e\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e\u003c/sup\u003e outlined the essential criteria for an effective flap perfusion monitoring system, requiring it to be accurate, reliable, applicable to all free flaps, easy for medical personnel to use, safe for both the patient and the flap, and capable of early detection. Swartz\u003csup\u003e\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e\u003c/sup\u003e further argued that an ideal monitoring method should provide continuous record patency and flow at the anastomosis, distinguish between arterial and venous compromise, and be applicable to both cutaneous and buried flaps.\u003c/p\u003e\u003cp\u003eA consistent feature across all free flap procedures, regardless of the presence of a skin paddle (the portion of skin included with the flap from the donor site), is the subcutaneous location of the pedicle (the nourishing vessels) and the relatively limited variation in its vessel caliber\u003csup\u003e\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e\u003c/sup\u003e (diameter of blood vessel). Our experimental design focused on developing a system that non-invasively accesses this pedicle, allowing for continuous measurement, discrimination between arterial and venous issues, and usability in buried flaps (transferred without a skin paddle, such as a muscle or bone flap, which is then covered by the native skin and is not visible on the surface).\u003c/p\u003e\u003cp\u003eThe distinctive temporal window of blood flow compromise,with 80% of thrombosis occurring within two days post-surgery, 90% of arterial thrombosis within 24 hours\u003csup\u003e\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u003c/sup\u003e, and salvage rates dropping significantly after five hours\u003csup\u003e\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e,is rarely seen in other disease groups. Developing a dedicated device focused solely on these tight parameters is commercially challenging due to the small market size, thereby stifling innovation. Consequently, achieving \"commercially available\" necessitates the application of established technological principles. In this context, PPG represents one of the most suitable technologies.\u003c/p\u003e\u003cp\u003eIn this study, silicone vascular phantoms were first fabricated. To evaluate their mechanical properties, Young\u0026rsquo;s modulus measurement and internal pressure testing were commonly considered. Internal pressure testing was chosen because it directly assessed compliance and the pressure\u0026ndash;diameter relationship under conditions similar to physiological blood flow\u003csup\u003e\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e\u003c/sup\u003e. In the internal pressure testing performed in this study, circumferential compliance was calculated, while longitudinal strain was not considered. This was because preliminary tests showed that when an internal pressure of 30 mmHg was applied to a soft silicone tube with a deadener ratio of 0.6, the axial elongation was only approximately 1%. The relationship between circumferential strain (εθ) and longitudinal strain (εz) can be expressed using Poisson\u0026rsquo;s ratio (ν) as follows\u003csup\u003e\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u003c/sup\u003e:\u003cdiv id=\"Equb\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\:\\epsilon\\:\\theta\\:=\\:\\epsilon\\:z\\:\\bullet\\:2\\:\\bullet\\:\\frac{1-v/2}{1-2v}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eSince ν is approximately 0.49 for soft silicone, εθ was calculated to be about 76 times greater than εz. Therefore, longitudinal strain was considered negligible.\u003c/p\u003e\u003cp\u003eThe pressure range of 40\u0026ndash;80 mmHg was set to simulate the internal thoracic artery, and 60\u0026ndash;120 mmHg was used to simulate the radial artery. As expected, increasing the deadener ratio increased both the change in cross-sectional area (ΔCSA) and compliance, whereas increasing the hardener ratio had the opposite effect. The arterial model showed the best agreement with physiological artery behaviour at a hardener ratio of H\u0026thinsp;=\u0026thinsp;0.6, where ΔCSA was 6.62\u0026ndash;8.09 mm\u0026sup2; (+\u0026thinsp;22.5%) at 40\u0026ndash;80 mmHg and 7.34\u0026ndash;9.14 mm\u0026sup2; (+\u0026thinsp;24.4%) at 60\u0026ndash;120 mmHg, which falls within reported values for the radial artery (3.4\u0026ndash;4.2 mm\u0026sup2;, +\u0026thinsp;23.5%) and brachial artery (8.1\u0026ndash;10.2 mm\u0026sup2;, +\u0026thinsp;25.9%)\u003csup\u003e32\u003c/sup\u003e. The compliance of the arterial model (2.99 \u0026times; 10⁻⁴ cm\u0026sup2;/mmHg) was also comparable to reported physiological values for the radial (1.3 \u0026times; 10⁻⁴ cm\u0026sup2;/mmHg) and brachial arteries (3.52 \u0026times; 10⁻⁴ cm\u0026sup2;/mmHg) \u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. In terms of scale, the cross-sectional area also matched the reported range for the deep inferior epigastric artery (3.14\u0026ndash;9.62 mm\u0026sup2;)\u003csup\u003e60\u003c/sup\u003e. Similarly, the venous model best reproduced physiological venous mechanics at a deadener ratio of D\u0026thinsp;=\u0026thinsp;0.4, where ΔCSA was 9.36\u0026ndash;12.5 mm\u0026sup2; (+\u0026thinsp;33.6%) at 10\u0026ndash;30 mmHg, which lies between reported values for the cephalic vein (5.5\u0026ndash;7.8 mm\u0026sup2;, +\u0026thinsp;41.8%) and basilic vein (16.9\u0026ndash;21.5 mm\u0026sup2;, +\u0026thinsp;27.2%)\u003csup\u003e32\u003c/sup\u003e. The compliance of the venous model (15.7 \u0026times; 10⁻⁴ cm\u0026sup2;/mmHg) was also consistent with reported venous values for the cephalic (5.6 \u0026times; 10⁻⁴ cm\u0026sup2;/mmHg) and basilic veins (17.4 \u0026times; 10⁻⁴ cm\u0026sup2;/mmHg) \u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u003c/sup\u003e. In terms of scale, the cross-sectional area was consistent with the reported range for the deep inferior epigastric vein (0.20\u0026ndash;14.50 mm\u0026sup2;)\u003csup\u003e60\u003c/sup\u003e. These arteries and veins were embedded in a tissue phantom whose silicone mixture ratio (A:B:D\u0026thinsp;=\u0026thinsp;1:1:1) was formulated to achieve a hardness corresponding to that of the finger and thigh\u003csup\u003e\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e,\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eThe internal pressure settings used for simulating the initial ischemic and congested pedicle states were established based on consultation with researchers who have performed continuous arterial and venous catheter-based pressure monitoring in over 200 free flap cases\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e. These settings primarily model pressures relevant to breast reconstruction; more pronounced pressure variations would be anticipated in reconstructions of the head and neck or extremities.\u003c/p\u003e\u003cp\u003eOur preprocessing methodology utilised literature-validated standard approaches, including the 4th-order Chebyshev Type II bandpass filter\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e, MAD-based artifact detection\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e, and cubic spline interpolation for baseline correction\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e. However, we acknowledge that no standardised preprocessing pipeline for PPG signals currently exists, with optimal cutoff frequencies varying significantly across studies\u003csup\u003e\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e\u003c/sup\u003e. While the literature notes that optimal filtering parameters vary across \u003cem\u003ein vivo\u003c/em\u003e subjects and conditions\u003csup\u003e\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e\u003c/sup\u003e, our approach employs a fixed bandpass filter (0.5\u0026ndash;10 Hz)\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e designed to cover the physiologically relevant heart rate range. Given the highly controlled, stationary nature of our \u003cem\u003ein vitro\u003c/em\u003e phantom and perfusion system, the established, fixed-parameter preprocessing pipeline ensures consistent and reliable signal conditioning across all experimental states.\u003c/p\u003e\u003cp\u003eDespite the validity of our literature-supported fixed approach, this limits our study. Future research should implement signal-specific adaptive optimisation. Such personalised filtering would offer substantial benefits for achieving higher precision in challenging measurement scenarios, such as the deep tissue measurements (21 mm) and pathological states (ischemia, congestion) investigated in this study.\u003c/p\u003e\u003cp\u003eThe evaluation of the Receiver Operating Characteristic (ROC) curves provides direct insight into the diagnostic capability of PPG features for distinguishing between vascular states. The results indicate that PPG features can separate free flap compromise based on distinct morphological characteristics.\u003c/p\u003e\u003cp\u003eIn the discrimination of Normal vs. Ischaemia, which represents a flow-occlusion anomaly marked by a drop in arterial pressure (AP), the IR Amplitude and IR Area features (e.g., IR_S_AUC, IR_D_AUC) achieved extremely high AUC values, ranging from 0.95 to 1.00. This observation demonstrates that Ischaemia directly causes a reduction in AP-dependent PPG amplitude and vascular volume. Specifically, the near-infrared channel accurately quantifies the absolute decrease in deep vessel blood volume, a primary indicator of ischemia, with near-perfect accuracy (AUC 1.00 at 15 mm). The high AUC values for diastolic area (D_AUC, 0.80\u0026ndash;0.97) are consistent with the corresponding reduction in diastolic vascular volume due to reduced arterial inflow. Time-based features like IR_Systolic Width (0.90) and IR_RiseTime (0.90) were also effective but less discriminative than the amplitude/volume metrics.\u003c/p\u003e\u003cp\u003eFor Normal vs. Congestion, an anomaly marked by elevated venous pressure (VP) and restricted outflow, the results were similar: IR Amplitude (0.96\u0026ndash;0.99) and Area features (0.87\u0026ndash;0.99) showed strong discriminatory power. Congestion causes blood to pool in the veins, leading to an increased vascular baseline (DC component) and suppression of the arterial pulsation amplitude (AC component). This confirms that PPG accurately captures the physical damping of the pulse wave due to venous stasis. The high accuracy of time-based features, such as IR_FallTime (0.94 at 3 mm) and IR_SlopeRatio (0.93), also confirms that the physical waveform elongation caused by delayed outflow is a simultaneous and detectable morphological marker.\u003c/p\u003e\u003cp\u003eThe comparison of Ischaemia vs. Congestion proved to be the most complex, as absolute amplitude is not sufficient for distinction. Discrimination was primarily driven by time and gradient features. IR_FallTime (0.97 at 3 mm) was particularly effective, differentiating the conditions based on the contrast between the prolonged outflow period in Congestion and the potentially faster pulse decay observed during Ischaemia due to low AP. Furthermore, the UpslopeLength_norm feature achieved an AUC of 1.00 at 9 mm, suggesting a complete difference in the complexity of the systolic rise profile between the two abnormalities. The relative strength of the Red/IR_Amplitude_ratio (0.78 at 21 mm) in deep tissue suggests that the physical difference in penetration depth allows the two wavelengths to capture complementary information about blood flow changes at varying tissue layers, which aids in differential diagnosis at depth.\u003c/p\u003e\u003cp\u003eIn summary, PPG features effectively classify free flap vascular compromise from a morphological perspective. The distinction of Ischaemia/Congestion vs. Normal states is predominantly achieved through IR Amplitude and Area features (0.96\u0026ndash;1.00), signifying that absolute vascular volume change is the core discriminator. In contrast, the vital discrimination between Ischaemia vs. Congestion relies on time and shape features (e.g., FallTime, Width Ratio), demonstrating that PPG accurately isolates the distinct physiological mechanisms of arterial pump failure versus venous outflow obstruction with high accuracy.\u003c/p\u003e\u003cp\u003eThis study employed phantom experiments, rather than human or animal clinical trials, to validate the theoretical diagnostic accuracy of Photoplethysmography (PPG) for early ischemia and early congestion in free flaps. Given that the incidence of vascular compromise in humans is low, ethical induction is difficult, and animal models present a scale mismatch, this research provides valuable insights.\u003c/p\u003e\u003cp\u003eA limitation of this study is the technical inability to perfectly replicate the dermis and subcutaneous tissue overlying the pedicle. Changes in blood flow within these superficial tissues could influence the measured values. However, if measurements are taken via the flap's skin paddle, this change in measured value might actually accentuate the AUC for differentiation. Future human clinical trials are required to clarify this relationship.\u003c/p\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eWith recent advancements in PPG technology, an \u003cem\u003ein vitro\u003c/em\u003e setting that can simulate normal, ischaemic, and congested states would be a beneficial tool for studies in free flap monitoring.\u003c/p\u003e\u003cp\u003eThe purposes of this study were twofold: 1) to develop a custom-made silicone phantom that simulates a free flap and to replicate ischaemic and congested states, and 2) to assess the capability to diagnose early flap disorder.\u003c/p\u003e\u003cp\u003ePPG signals obtained from this model were recorded and analysed, allowing us to detect the morphological changes caused by each condition. A statistically significant change was detected, which demonstrates the potential of PPG technology for detecting flap blood flow impairment.\u003c/p\u003e\u003cp\u003eGoing forward, work should also aim to further improve the similarity of the phantom to human tissue by exploring different materials and generating signals that are even closer to those found in the human body. Furthermore, it is necessary to investigate more advanced signal feature extraction and highly accurate anomaly detection methods.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cb\u003eFunding Declaration\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThis research was supported by the Pump Priming Funds for consumables awarded by the British Association of Plastic, Reconstructive and Aesthetic Surgeons (BAPRAS). The authors would like to express their sincere gratitude to BAPRAS for the financial support that made this work possible.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eHiroki Kodama: Conceptualization, Investigation, Formal analysis, Writing -original draft, Project administration, Methodology, Visualization, Data curationJames May: Conceptualization, Writing -reviewing and editingDariush Nikkhah: Writing -reviewing and editingPanicos Kyriacou: Writing -reviewing\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eI sincerely thank Dr. Saiga for his profound knowledge and expert guidance on setting the precise internal arterial and venous pressure parameters for the in vitro experiments, drawing from his extensive experience with over 200 cases of continuous catheter-based pressure monitoring in free flaps. I also wish to thank my laboratory colleagues for laying the foundational work of this research and my colleagues who assisted with the waveform quality assessment.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eThe datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eNakatsuka, T. et al. Analytic Review of 2372 Free Flap Transfers for Head and Neck Reconstruction Following Cancer Resection. \u003cem\u003eJ. reconstr. Microsurg\u003c/em\u003e. \u003cb\u003e19\u003c/b\u003e (06), 363\u0026ndash;368. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1055/s-2003-42630\u003c/span\u003e\u003cspan address=\"10.1055/s-2003-42630\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2003).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHarashina, T., Imai, T., Nakajima, H. \u0026amp; Fujino, T. Breast reconstruction with microsurgical free composite tissue transplantation. \u003cem\u003eBr. J. Plast. Surg.\u003c/em\u003e \u003cb\u003e33\u003c/b\u003e (1), 30\u0026ndash;37. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/0007-1226(80)90048-X\u003c/span\u003e\u003cspan address=\"10.1016/0007-1226(80)90048-X\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1980).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTaylor, G. I., Miller, G. D. H., Ham, F. J. \u0026amp; THE FREE VASCULARIZED BONE GRAFT: A Clinical Extension of Microvascular Techniques. \u003cem\u003ePlast. Reconstr. Surg.\u003c/em\u003e \u003cb\u003e55\u003c/b\u003e (5), 533\u0026ndash;544. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/00006534-197505000-00002\u003c/span\u003e\u003cspan address=\"10.1097/00006534-197505000-00002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1975).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003e2020 Plastic Surgery Statistics Report. American Society of Plastic Surgeons Accessed October 13. (2025). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.plasticsurgery.org/documents/news/statistics/2020/plastic-surgery-statistics-full-report-2020.pdf\u003c/span\u003e\u003cspan address=\"https://www.plasticsurgery.org/documents/news/statistics/2020/plastic-surgery-statistics-full-report-2020.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCaine, P., Jeevaratnam, J. A., Misky, A. \u0026amp; Nikkhah, D. Post-operative Monitoring. In: Core Techniques in Flap Reconstructive Microsurgery - A Stepwise Guide. Springer Nature Switzerland AG; :39\u0026ndash;46. (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTran, N. V., Buchel, E. W. \u0026amp; Convery, P. A. Microvascular Complications of DIEP Flaps. \u003cem\u003ePlast. Reconstr. Surg.\u003c/em\u003e \u003cb\u003e119\u003c/b\u003e (5), 1397\u0026ndash;1405. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/01.prs.0000256045.71765.96\u003c/span\u003e\u003cspan address=\"10.1097/01.prs.0000256045.71765.96\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2007).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBui, D. T. et al. Free Flap Reexploration: Indications, Treatment, and Outcomes in 1193 Free Flaps. \u003cem\u003ePlast. Reconstr. Surg.\u003c/em\u003e \u003cb\u003e119\u003c/b\u003e (7), 2092\u0026ndash;2100. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/01.prs.0000260598.24376.e1\u003c/span\u003e\u003cspan address=\"10.1097/01.prs.0000260598.24376.e1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2007).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKroll, S. S. et al. Timing of Pedicle Thrombosis and Flap Loss after Free-Tissue Transfer. \u003cem\u003ePlast. Reconstr. Surg.\u003c/em\u003e \u003cb\u003e98\u003c/b\u003e (7), 1230\u0026ndash;1233. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/00006534-199612000-00017\u003c/span\u003e\u003cspan address=\"10.1097/00006534-199612000-00017\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1996).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMericli, A. F. et al. A Prospective Clinical Trial Comparing Visible Light Spectroscopy to Handheld Doppler for Postoperative Free Tissue Transfer Monitoring. \u003cem\u003ePlast. Reconstr. Surg.\u003c/em\u003e \u003cb\u003e140\u003c/b\u003e (3), 604\u0026ndash;613. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/PRS.0000000000003600\u003c/span\u003e\u003cspan address=\"10.1097/PRS.0000000000003600\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDisa, J. J., Cordeiro, P. G. \u0026amp; Hidalgo, D. A. Efficacy of conventional monitoring techniques in free tissue transfer: an 11-year experience in 750 consecutive cases. \u003cem\u003ePlast. Reconstr. Surg.\u003c/em\u003e \u003cb\u003e104\u003c/b\u003e (1), 97\u0026ndash;101 (1999).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen, K. T. et al. Timing of Presentation of the First Signs of Vascular Compromise Dictates the Salvage Outcome of Free Flap Transfers. \u003cem\u003ePlast. Reconstr. Surg.\u003c/em\u003e \u003cb\u003e120\u003c/b\u003e (1), 187\u0026ndash;195. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/01.prs.0000264077.07779.50\u003c/span\u003e\u003cspan address=\"10.1097/01.prs.0000264077.07779.50\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2007).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAl-Dam, A. et al. Outcome of microvascular free flaps in a high-volume training centre. \u003cem\u003eJ. Cranio-Maxillofacial Surg.\u003c/em\u003e \u003cb\u003e42\u003c/b\u003e (7), 1178\u0026ndash;1183. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jcms.2014.02.005\u003c/span\u003e\u003cspan address=\"10.1016/j.jcms.2014.02.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2014).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePatel, U. A. et al. Free Flap Reconstruction Monitoring Techniques and Frequency in the Era of Restricted Resident Work Hours. \u003cem\u003eJAMA Otolaryngol. Head Neck Surg.\u003c/em\u003e \u003cb\u003e143\u003c/b\u003e (8), 803. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1001/jamaoto.2017.0304\u003c/span\u003e\u003cspan address=\"10.1001/jamaoto.2017.0304\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKagaya, Y. \u0026amp; Miyamoto, S. A systematic review of near-infrared spectroscopy in flap monitoring: Current basic and clinical evidence and prospects. \u003cem\u003eJ. Plast. Reconstr. Aesthetic Surg.\u003c/em\u003e \u003cb\u003e71\u003c/b\u003e (2), 246\u0026ndash;257. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.bjps.2017.10.020\u003c/span\u003e\u003cspan address=\"10.1016/j.bjps.2017.10.020\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKodama, H. et al. The future of free flap monitoring by laser continuous doppler flowmetry: A prospective assessment in consecutive 71 patients. \u003cem\u003eJPRAS Open.\u003c/em\u003e \u003cb\u003e43\u003c/b\u003e, 140\u0026ndash;152. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jpra.2024.11.004\u003c/span\u003e\u003cspan address=\"10.1016/j.jpra.2024.11.004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWu, K. G. et al. Implantable arterial doppler efficacy in free flap reconstruction: A systematic review and meta-analysis. \u003cem\u003eHead \u0026amp; Neck\u003c/em\u003e. Published online August 29, :hed.27500. (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/hed.27500\u003c/span\u003e\u003cspan address=\"10.1002/hed.27500\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eColakoglu, S., Johnson, A., Anderson, J., Mathes, D. W. \u0026amp; Chong, T. W. Changes to Venous Flow Coupler Signal during DIEP Flap Inset Can Be Predictive of Poor Clinical Outcomes in Autologous Breast Reconstruction. \u003cem\u003eJ. Reconstr. Microsurg\u003c/em\u003e. \u003cb\u003e36\u003c/b\u003e (6), 466\u0026ndash;470. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1055/s-0040-1703014\u003c/span\u003e\u003cspan address=\"10.1055/s-0040-1703014\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHwang, S. J. Diagnostic Accuracy of Microdialysis in Postoperative Flap Monitoring. \u003cem\u003eJ. Craniofac. Surg.\u003c/em\u003e \u003cb\u003e34\u003c/b\u003e (1), 288\u0026ndash;290. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/SCS.0000000000008878\u003c/span\u003e\u003cspan address=\"10.1097/SCS.0000000000008878\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBurnier, P., Niddam, J., Bosc, R., Hersant, B. \u0026amp; Meningaud, J. P. Indocyanine green applications in plastic surgery: A review of the literature. \u003cem\u003eJ. Plast. Reconstr. Aesthetic Surg.\u003c/em\u003e \u003cb\u003e70\u003c/b\u003e (6), 814\u0026ndash;827. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.bjps.2017.01.020\u003c/span\u003e\u003cspan address=\"10.1016/j.bjps.2017.01.020\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2017).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHudson, T., Hogue, E., Mullner, D., Herrera, F. \u0026amp; Scomacao, I. The Utility of Smartphone-Based Thermal Imaging in the Management and Monitoring of Microvascular Flap Procedures: A Systematic Review and Meta-Analysis. \u003cem\u003eAnn. Plast. Surg.\u003c/em\u003e \u003cb\u003e90\u003c/b\u003e (6S Suppl 4), S420\u0026ndash;S425. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/SAP.0000000000003343\u003c/span\u003e\u003cspan address=\"10.1097/SAP.0000000000003343\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKwasnicki, R. M., Noakes, A. J., Banhidy, N. \u0026amp; Hettiaratchy, S. Quantifying the Limitations of Clinical and Technology-based Flap Monitoring Strategies using a Systematic Thematic Analysis. \u003cem\u003ePlast. Reconstr. Surg. - Global Open.\u003c/em\u003e \u003cb\u003e9\u003c/b\u003e (7), e3663. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/GOX.0000000000003663\u003c/span\u003e\u003cspan address=\"10.1097/GOX.0000000000003663\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2021).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKyriacou, P. A. \u003cem\u003ePhotoplethysmography: Technology, Signal Analysis and Applications\u003c/em\u003e (Elsevier, AP Academic, 2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCharlton, P. H. et al. The 2023 wearable photoplethysmography roadmap. \u003cem\u003ePhysiol. Meas.\u003c/em\u003e \u003cb\u003e44\u003c/b\u003e (11), 111001. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1088/1361-6579/acead2\u003c/span\u003e\u003cspan address=\"10.1088/1361-6579/acead2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZaman, T., Kyriacou, P. A. \u0026amp; Pal, S. K. Development of a reflectance photoplethysmographic sensor used for the assessment of free flap perfusion. In: \u003cem\u003e2011 Annual International Conference of the IEEE Engineering in Medicine and Biology Society\u003c/em\u003e. IEEE; :4006\u0026ndash;4009. (2011). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/IEMBS.2011.6090995\u003c/span\u003e\u003cspan address=\"10.1109/IEMBS.2011.6090995\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKyriacou, P. A., Zaman, T. \u0026amp; Pal, S. K. Photoplethysmography in postoperative monitoring of deep inferior epigastric perforator (DIEP) free flaps. \u003cem\u003ePhysiol. Meas.\u003c/em\u003e \u003cb\u003e41\u003c/b\u003e (12), 124001. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1088/1361-6579/abc4c8\u003c/span\u003e\u003cspan address=\"10.1088/1361-6579/abc4c8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHong, K. S., Park, K. T. \u0026amp; Ahn, J. M. Aging Index using Photoplethysmography for a Healthcare Device: Comparison with Brachial-Ankle Pulse Wave Velocity. \u003cem\u003eHealthc. Inf. Res.\u003c/em\u003e \u003cb\u003e21\u003c/b\u003e (1), 30. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.4258/hir.2015.21.1.30\u003c/span\u003e\u003cspan address=\"10.4258/hir.2015.21.1.30\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2015).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGrabovskis, A., Marcinkevics, Z., Rubins, U. \u0026amp; Kviesis-Kipge, E. Effect of probe contact pressure on the photoplethysmographic assessment of conduit artery stiffness. \u003cem\u003eJ. Biomed. Opt.\u003c/em\u003e \u003cb\u003e18\u003c/b\u003e (2), 027004. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1117/1.JBO.18.2.027004\u003c/span\u003e\u003cspan address=\"10.1117/1.JBO.18.2.027004\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMillasseau, S. C., Ritter, J. M., Takazawa, K. \u0026amp; Chowienczyk, P. J. Contour analysis of the photoplethysmographic pulse measured at the finger. \u003cem\u003eJ. Hypertens.\u003c/em\u003e \u003cb\u003e24\u003c/b\u003e (8), 1449\u0026ndash;1456. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/01.hjh.0000239277.05068.87\u003c/span\u003e\u003cspan address=\"10.1097/01.hjh.0000239277.05068.87\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2006).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBudidha, K., Rybynok, V. \u0026amp; Kyriacou, P. A. Design and Development of a Modular, Multichannel Photoplethysmography System. \u003cem\u003eIEEE Trans. Instrum. Meas.\u003c/em\u003e \u003cb\u003e67\u003c/b\u003e (8), 1954\u0026ndash;1965. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/TIM.2018.2810643\u003c/span\u003e\u003cspan address=\"10.1109/TIM.2018.2810643\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFerizoli, R., Karimpour, P., May, J. M. \u0026amp; Kyriacou, P. A. Arterial stiffness assessment using PPG feature extraction and significance testing in an in vitro cardiovascular system. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e14\u003c/b\u003e (1), 2024. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-024-51395-y\u003c/span\u003e\u003cspan address=\"10.1038/s41598-024-51395-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKarimpour, P., Ferizoli, R., May, J. M. \u0026amp; Kyriacou, P. A. Customisable Silicone Vessels and Tissue Phantoms for In Vitro Photoplethysmography Investigations into Cardiovascular Disease. \u003cem\u003eSensors\u003c/em\u003e \u003cb\u003e24\u003c/b\u003e (5), 1681. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/s24051681\u003c/span\u003e\u003cspan address=\"10.3390/s24051681\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCappelletti, S. et al. Non-invasive estimation of vascular compliance and distensibility in the arm vessels: a novel ultrasound-based protocol. \u003cem\u003eQuant. Imaging Med. Surg.\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e (7), 3515\u0026ndash;3527. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.21037/qims-21-987\u003c/span\u003e\u003cspan address=\"10.21037/qims-21-987\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNomoni, M., May, J. M. \u0026amp; Kyriacou, P. A. Novel Polydimethylsiloxane (PDMS) Pulsatile Vascular Tissue Phantoms for the In-Vitro Investigation of Light Tissue Interaction in Photoplethysmography. \u003cem\u003eSensors\u003c/em\u003e \u003cb\u003e20\u003c/b\u003e (15), 4246. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/s20154246\u003c/span\u003e\u003cspan address=\"10.3390/s20154246\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChan, G. et al. Multi-Site Photoplethysmography Technology for Blood Pressure Assessment: Challenges and Recommendations. \u003cem\u003eJCM\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e (11), 1827. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/jcm8111827\u003c/span\u003e\u003cspan address=\"10.3390/jcm8111827\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2019).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFerizoli, R., Karimpour, P., May, J. M. \u0026amp; Kyriacou, P. A. A Bilateral In Vitro Model for Cardiovascular Disease Investigations Using Photoplethysmography Sensors. In: \u003cem\u003e2023 IEEE BioSensors Conference (BioSensors)\u003c/em\u003e. IEEE; :1\u0026ndash;4. (2023). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/BioSensors58001.2023.10281067\u003c/span\u003e\u003cspan address=\"10.1109/BioSensors58001.2023.10281067\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eUV-Vis Spectrum of Methylene Blue Solution, with Maximu|Open-i. Available online: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://openi.nlm.nih.gov/detailedresult?img=PMC2808005_ijms-10-04707f6\u0026amp;req=4\u003c/span\u003e\u003cspan address=\"https://openi.nlm.nih.gov/detailedresult?img=PMC2808005_ijms-10-04707f6\u0026amp;req=4\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (accessed on 16 January 2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMatsumine, H. et al. Improvement of the salvage-rate of flap after venous thrombosis with intraparenchymatous venous pressure monitoring. \u003cem\u003eMicrosurgery\u003c/em\u003e \u003cb\u003e38\u003c/b\u003e (5), 498\u0026ndash;503. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/micr.30293\u003c/span\u003e\u003cspan address=\"10.1002/micr.30293\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSaiga, A., Kubota, Y., Yamaji, Y. \u0026amp; Mitsukawa, N. Intraflap Vascular Catheterization Method for Monitoring, Prevention, and Intervention of Thrombogenesis in Free-Flap Surgery. \u003cem\u003eAnn. Plast. Surg.\u003c/em\u003e \u003cb\u003e88\u003c/b\u003e (1), 68\u0026ndash;73. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/SAP.0000000000003049\u003c/span\u003e\u003cspan address=\"10.1097/SAP.0000000000003049\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGoda, M. \u0026Aacute;., Charlton, P. H. \u0026amp; Behar, J. A. pyPPG: a Python toolbox for comprehensive photoplethysmography signal analysis. \u003cem\u003ePhysiol. Meas.\u003c/em\u003e \u003cb\u003e45\u003c/b\u003e (4), 045001. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1088/1361-6579/ad33a2\u003c/span\u003e\u003cspan address=\"10.1088/1361-6579/ad33a2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2024).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiang, Y., Chen, Z., Ward, R. \u0026amp; Elgendi, M. Photoplethysmography and Deep Learning: Enhancing Hypertension Risk Stratification. \u003cem\u003eBiosensors\u003c/em\u003e \u003cb\u003e8\u003c/b\u003e (4), 101. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/bios8040101\u003c/span\u003e\u003cspan address=\"10.3390/bios8040101\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eOppenheim, A. V. \u0026amp; Schafer, R. W. \u003cem\u003eDiscrete-Time Signal Processing\u003c/em\u003e (Pearson, 2010). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://books.google.co.uk/books?id=mYsoAQAAMAAJ\u003c/span\u003e\u003cspan address=\"https://books.google.co.uk/books?id=mYsoAQAAMAAJ\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLeys, C., Ley, C., Klein, O., Bernard, P. \u0026amp; Licata, L. Detecting outliers: Do not use standard deviation around the mean, use absolute deviation around the median. \u003cem\u003eJ. Exp. Soc. Psychol.\u003c/em\u003e \u003cb\u003e49\u003c/b\u003e (4), 764\u0026ndash;766. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jesp.2013.03.013\u003c/span\u003e\u003cspan address=\"10.1016/j.jesp.2013.03.013\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMiller, J. Short Report: Reaction Time Analysis with Outlier Exclusion: Bias Varies with Sample Size. \u003cem\u003eQ. J. Experimental Psychol. Sect. A\u003c/em\u003e. \u003cb\u003e43\u003c/b\u003e (4), 907\u0026ndash;912. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/14640749108400962\u003c/span\u003e\u003cspan address=\"10.1080/14640749108400962\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1991).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRam, M. R., Madhav, K. V., Krishna, E. H., Komalla, N. R. \u0026amp; Reddy, K. A. A Novel Approach for Motion Artifact Reduction in PPG Signals Based on AS-LMS Adaptive Filter. \u003cem\u003eIEEE Trans. Instrum. Meas.\u003c/em\u003e \u003cb\u003e61\u003c/b\u003e (5), 1445\u0026ndash;1457. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/TIM.2011.2175832\u003c/span\u003e\u003cspan address=\"10.1109/TIM.2011.2175832\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang, L., Lo, B. P. \u0026amp; Yang, G. Z. Multichannel Reflective PPG Earpiece Sensor With Passive Motion Cancellation. \u003cem\u003eIEEE Trans. Biomed. Circuits Syst.\u003c/em\u003e \u003cb\u003e1\u003c/b\u003e (4), 235\u0026ndash;241. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/TBCAS.2007.910900\u003c/span\u003e\u003cspan address=\"10.1109/TBCAS.2007.910900\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2007).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJang, D. G., Park, S., Hahn, M. \u0026amp; Park, S. H. A Real-Time Pulse Peak Detection Algorithm for the Photoplethysmogram. \u003cem\u003eIJEEE Published online\u003c/em\u003e \u003cb\u003e2014\u003c/b\u003e:45\u0026ndash;49. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.12720/ijeee.2.1.45-49\u003c/span\u003e\u003cspan address=\"10.12720/ijeee.2.1.45-49\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMillasseau, S. C., Ritter, J. M., Takazawa, K. \u0026amp; Chowienczyk, P. J. Contour analysis of the photoplethysmographic pulse measured at the finger. \u003cem\u003eJ. Hypertens.\u003c/em\u003e \u003cb\u003e24\u003c/b\u003e (8), 1449\u0026ndash;1456. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/01.hjh.0000239277.05068.87\u003c/span\u003e\u003cspan address=\"10.1097/01.hjh.0000239277.05068.87\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2006).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePang, J. et al. Nonparametric bootstrap methods for interval estimation of the area under the ROC curve with correlated diagnostic test data: application to whole-virus ELISA testing in swine. \u003cem\u003eFront. Vet. Sci.\u003c/em\u003e \u003cb\u003e10\u003c/b\u003e, 1274786. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3389/fvets.2023.1274786\u003c/span\u003e\u003cspan address=\"10.3389/fvets.2023.1274786\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFalanga, V. \u0026amp; Bucalo, B. Use of a durometer to assess skin hardness. \u003cem\u003eJ. Am. Acad. Dermatol.\u003c/em\u003e \u003cb\u003e29\u003c/b\u003e (1), 47\u0026ndash;51. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/0190-9622(93)70150-R\u003c/span\u003e\u003cspan address=\"10.1016/0190-9622(93)70150-R\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1993).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePereira, T. et al. A Supervised Approach to Robust Photoplethysmography Quality Assessment. \u003cem\u003eIEEE J. Biomed. Health Inf.\u003c/em\u003e \u003cb\u003e24\u003c/b\u003e (3), 649\u0026ndash;657. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/JBHI.2019.2909065\u003c/span\u003e\u003cspan address=\"10.1109/JBHI.2019.2909065\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2020).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCharlton, P. H., Marozas, V., Mej\u0026iacute;a-Mej\u0026iacute;a, E., Kyriacou, P. A. \u0026amp; Mant, J. Determinants of photoplethysmography signal quality at the wrist. \u003cem\u003ePLOS Digit. Health\u003c/em\u003e. \u003cb\u003e4\u003c/b\u003e (6), e0000585. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pdig.0000585\u003c/span\u003e\u003cspan address=\"10.1371/journal.pdig.0000585\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eReconstructive, B. A., Surgeons, P. \u0026amp; Surgery, A. of A of B. Oncoplastic Breast Reconstruction Guidelines for Best Practice. Published online 2012. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.bapras.org.uk/docs/default-source/commissioning-and-policy/final-oncoplastic-guidelines\u0026ndash;healthcare-professionals.pdf?sfvrsn=0\u003c/span\u003e\u003cspan address=\"http://www.bapras.org.uk/docs/default-source/commissioning-and-policy/final-oncoplastic-guidelines\u0026ndash;healthcare-professionals.pdf?sfvrsn=0\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCreech, B. \u0026amp; Miller, S. Evaluation of circulation in skin flaps. \u003cem\u003eSkin Flaps Boston: Little Brown\u003c/em\u003e ;\u003cb\u003e21\u003c/b\u003e. (1975).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSwartz, W. M., Jones, N. F., Cherup, L., Klein, A. \u0026amp; Pittsburgh Direct Monitoring of Microvascular Anastomoses with the 20-MHz Ultrasonic Doppler Probe: An Experimental and Clinical Study. \u003cem\u003ePlast. Reconstr. Surg.\u003c/em\u003e \u003cb\u003e81\u003c/b\u003e (2), 149\u0026ndash;158. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/00006534-198802000-00001\u003c/span\u003e\u003cspan address=\"10.1097/00006534-198802000-00001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1988).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGeoffrey, G. \u0026amp; Hallock Chapter 3: Flap selection. In: Flaps and Reconstructive Surgery. Elsevier Inc; :17. (2009).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTomioka, Y. et al. Wearable, wireless, multi-sensor device for monitoring tissue circulation after free-tissue transplantation: a multicentre clinical trial. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e12\u003c/b\u003e (1), 16532. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-022-21007-8\u003c/span\u003e\u003cspan address=\"10.1038/s41598-022-21007-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMin, K., Hong, J. P. \u0026amp; Suh, H. P. Risk Factors for Partial Flap Loss in a Free Flap: A 12-Year Retrospective Study of Anterolateral Thigh Free Flaps in 303 Lower Extremity Cases. \u003cem\u003ePlast. Reconstr. Surg.\u003c/em\u003e \u003cb\u003e150\u003c/b\u003e (5), 1071e\u0026ndash;1081e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/PRS.0000000000009646\u003c/span\u003e\u003cspan address=\"10.1097/PRS.0000000000009646\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2022).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMacrae, R. A., Miller, K. \u0026amp; Doyle, B. J. Methods in Mechanical Testing of Arterial Tissue: A Review. \u003cem\u003eStrain\u003c/em\u003e \u003cb\u003e52\u003c/b\u003e (5), 380\u0026ndash;399. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/str.12183\u003c/span\u003e\u003cspan address=\"10.1111/str.12183\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePajic-Lijakovic, I., Milivojevic, M. \u0026amp; McClintock, P. V. E. Compressibility of biological systems: the viscoelastic Poisson\u0026rsquo;s ratio. \u003cem\u003eAdv. Physics: X\u003c/em\u003e. \u003cb\u003e10\u003c/b\u003e (1), 2440023. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/23746149.2024.2440023\u003c/span\u003e\u003cspan address=\"10.1080/23746149.2024.2440023\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAyhan, S. et al. Correlation between vessel diameters of superficial and deep inferior epigastric systems: Doppler ultrasound assessment. \u003cem\u003eJ. Plast. Reconstr. Aesthetic Surg.\u003c/em\u003e \u003cb\u003e62\u003c/b\u003e (9), 1140\u0026ndash;1147. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.bjps.2008.02.012\u003c/span\u003e\u003cspan address=\"10.1016/j.bjps.2008.02.012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2009).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFalanga, V. \u0026amp; Bucalo, B. Use of a durometer to assess skin hardness. \u003cem\u003eJ. Am. Acad. Dermatol.\u003c/em\u003e \u003cb\u003e29\u003c/b\u003e (1), 47\u0026ndash;51. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/0190-9622(93)70150-R\u003c/span\u003e\u003cspan address=\"10.1016/0190-9622(93)70150-R\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (1993).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eOprişan, C. et al. Experimental determination of the Young\u0026rsquo;s modulus for the fingers with application in prehension systems for small cylindrical objects. \u003cem\u003eIOP Conf. Ser: Mater. Sci. Eng.\u003c/em\u003e \u003cb\u003e147\u003c/b\u003e, 012058. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1088/1757-899X/147/1/012058\u003c/span\u003e\u003cspan address=\"10.1088/1757-899X/147/1/012058\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2016).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWatanabe, Y., Yamane, N., Sathyanarayana, A., Mishra, V. \u0026amp; Goodwin, M. S. Beyond Motion Artifacts: Optimizing PPG Preprocessing for Accurate Pulse Rate Variability Estimation. \u003cem\u003earXiv\u003c/em\u003e. Preprint posted online 2025. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.48550/ARXIV.2510.06158\u003c/span\u003e\u003cspan address=\"10.48550/ARXIV.2510.06158\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSahan, B., Yolcu, C. \u0026amp; Sahin, S. Energy consumption and boot time performances of Kentkart industrial tablet on Android platform. In: \u003cem\u003eInternational Conference on Engineering and\u003c/em\u003e Technology \u003cem\u003e(ICET)\u003c/em\u003e. IEEE; 2017:1\u0026ndash;4. IEEE; 2017:1\u0026ndash;4. (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/ICEngTechnol.2017.8308177\u003c/span\u003e\u003cspan address=\"10.1109/ICEngTechnol.2017.8308177\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJoseph, G., Joseph, A., Titus, G., Thomas, R. M. \u0026amp; Jose, D. Photoplethysmogram (PPG) signal analysis and wavelet de-noising. In: \u003cem\u003e2014 Annual International Conference on Emerging Research Areas: Magnetics, Machines and Drives (AICERA/iCMMD)\u003c/em\u003e. IEEE; :1\u0026ndash;5. (2014). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/AICERA.2014.6908199\u003c/span\u003e\u003cspan address=\"10.1109/AICERA.2014.6908199\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Free Flap, Photoplethysmography (PPG), Phantom study, Plastic Surgery, Biomedical Engineering, Wearable Device","lastPublishedDoi":"10.21203/rs.3.rs-8080454/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8080454/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e\u003cp\u003eFree flap reconstruction is a cornerstone of plastic surgery, yet vascular thrombosis, a blockage of the supporting blood vessels,(3%\u0026ndash;5% incidence) requires early detection for salvage. Traditional clinical assessment is subjective and labour-intensive, driving the need for reliable monitoring. Photoplethysmography (PPG), a non-invasive, continuous optical sensing technology monitoring vascular volume, is highly suited for this purpose.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e\u003cp\u003eThis study investigated PPG's diagnostic capability for early vascular changes utilising a custom silicone phantom as a test-bed. Vessels were fabricated to mimic human arterial and venous mechanics (compliance, ΔCSA) and embedded at depths of 3,9,15, and 21 mm. A perfusion system simulated normal, early ischemic (low arterial pressure), and early congested (high venous pressure) states. Signals were analysed using 34 pulse-wave features.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e\u003cp\u003eAnalysis of the diagnostic performance (AUC\u0026thinsp;\u0026gt;\u0026thinsp;0.70) showed that PPG (IR) could successfully diagnose ischemia up to a depth of 21 mm and congestion up to 15 mm. The system was able to successfully distinguish between ischemia and congestion at all depths.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e\u003cp\u003eThis research confirms that specific PPG features are discernible indicators of early blood flow changes. The high diagnostic accuracy achieved in this \u003cem\u003ein vitro\u003c/em\u003e model establishes PPG as a robust platform for developing future continuous and objective free flap monitoring systems.\u003c/p\u003e","manuscriptTitle":"Diagnostic Performance of Photoplethysmography (PPG) for Early Vascular Compromise in a Customisable In Vitro Flap Model","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-11-26 16:33:41","doi":"10.21203/rs.3.rs-8080454/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"29b85995-d7db-49f3-94b5-59a9c1432aa0","owner":[],"postedDate":"November 26th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":58535220,"name":"Health sciences/Diseases"},{"id":58535221,"name":"Physical sciences/Engineering"},{"id":58535222,"name":"Health sciences/Health care"},{"id":58535223,"name":"Health sciences/Medical research"}],"tags":[],"updatedAt":"2026-02-23T07:10:38+00:00","versionOfRecord":[],"versionCreatedAt":"2025-11-26 16:33:41","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8080454","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8080454","identity":"rs-8080454","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}