Unsupervised Identification of Intraoperative Cardiopulmonary Interaction States During General Anesthesia

Unsupervised Identification of Intraoperative Cardiopulmonary Interaction States During General Anesthesia | 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 Research Article Unsupervised Identification of Intraoperative Cardiopulmonary Interaction States During General Anesthesia Albion Dervishi This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8961496/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 5 You are reading this latest preprint version Abstract Background Intraoperative hemodynamic instability may result from disruption of heart-lung interaction rather than isolated abnormalities in individual variables. We evaluated whether unsupervised analysis of routine intraoperative monitoring data can identify physiologically coherent cardiopulmonary interaction states. Methods Intraoperative data from 70 adults (21,510 synchronized observations) in the VitalDB database were analyzed. An unsupervised analytical framework was applied to multivariate physiological signals. Identified deviations were stratified by concurrent cardiac output to define physiological interaction regimes. Feature importance and correlation analyses were used to characterize physiological patterns. Results Three interaction regimes emerged: coupled physiology, low-flow/high-resistance, and hyperdynamic/low-resistance states. The low-flow state demonstrated reduced stroke volume with elevated systemic vascular resistance, consistent with vasoconstrictive compensation. The hyperdynamic state was characterized by increased stroke volume, reduced resistance, tachycardia, and lower diastolic pressure. Ventilatory-hemodynamic relationships were state dependent, revealing nonlinear tipping behavior under rising ventilatory load. Interaction states were identifiable using routine monitoring signals without reliance on invasive cardiac output measurement. Conclusions Cardiopulmonary interaction states represent multivariate, state-dependent physiological patterns detectable through unsupervised analysis of routine intraoperative data. Cardiopulmonary interaction Hemodynamic instability General anesthesia Mechanical ventilation Intraoperative monitoring Physiological modeling Figures Figure 1 Figure 2 Figure 3 Figure 4 1. Introduction 1.1 Clinical significance of cardiopulmonary interaction failure Intraoperative hemodynamic instability is a frequent and clinically significant challenge during general anesthesia, particularly among patients with limited cardiovascular reserve ( 1 ). Unrecognized rapid development of low-cardiac-output states may result in tissue hypoperfusion, postoperative organ dysfunction, prolonged intensive care unit stay( 2 ), and increased mortality. Early identification of preload-dependent states, such as hypovolemia or vasodilated conditions, and their associated cardiopulmonary interactions is critical for preventing circulatory collapse. This need may be particularly relevant during trauma, ongoing hemorrhage, or anesthetic induction, when physiological reserve is limited ( 3 ). 1.2 Heart-Lung Interaction as a State Variable Identical ventilatory input can increase cardiac output in high-resistance states by unloading the left ventricle. However, in preload-dependent or vasodilated states, the same input may reduce cardiac output by limiting venous return( 3 – 7 ). 1.3 Physiological meaning emerges from signal interactions Intraoperative signals, such as arterial pressure, oxygenation, end-tidal CO₂, ventilatory parameters, and anesthetic dosing, are continuously recorded but are often interpreted as isolated variables. Clinically meaningful deterioration may arise from altered relationships among these signals rather than from threshold violations of individual parameters( 8 ). This perspective supports a framework in which physiological normality and deviation are defined by patterns of interaction among signals, rather than by isolated measurements( 9 , 10 ). 1.4 Unsupervised identification of interaction states This study evaluated whether unsupervised analysis of multivariate intraoperative data can identify coherent cardiopulmonary interaction states. An unsupervised analytical framework was applied to detect deviations from the prevailing multivariate physiological structure, thereby revealing altered coupling between ventilatory load and hemodynamic response( 11 ). Unlike supervised models that predict predefined outcomes, this approach seeks to uncover intrinsic interaction regimes within routine monitoring data. The objective is not to replace clinical judgment but to provide a physiology-based method for characterizing cardiopulmonary interaction states grounded in established cardiovascular and pulmonary mechanisms. 2. Materials and Methods 2.1 Study Design and Data Source This retrospective analysis utilized de-identified intraoperative physiological data from the VitalDB open-access database, which contains synchronized recordings from patients who underwent non-cardiac surgery at Seoul National University Hospital between August 2016 and June 2017( 12 ). Data collection was conducted prospectively with institutional review board approval (IRB No. H-1408-101-605) and are publicly available (NCT02914444). The dataset comprises continuous waveform and numeric recordings from multiple perioperative monitoring systems across 6,388 surgical cases. 2.2 Patient selection Adult patients were included if continuous intraoperative recordings were available from patient monitors, anesthesia machines, bispectral index monitoring, infusion pumps, and advanced hemodynamic monitoring (EV1000). To reduce confounding from major hemorrhage, cases with estimated blood loss ≥ 500 mL were excluded. After applying these criteria, 70 patients were included in the final analysis. 2.3 Data Acquisition and Preprocessing The analysis included six static variables (age, sex, height, weight, preoperative hemoglobin, and albumin) and 23 intraoperative dynamic parameters collected from multiple monitoring systems. These parameters comprised invasive arterial pressure, heart rate, oxygen saturation, body temperature, ventilatory variables, anesthetic infusion rates, depth of anesthesia (BIS), and advanced hemodynamic variables such as cardiac output, cardiac index, stroke volume, systemic vascular resistance, and central venous pressure. Due to varying sampling frequencies across devices, only time-synchronized observations were retained. Implausible negative values and artefactual readings were excluded from the dataset. Body temperature values below 33°C were classified as sensor errors and removed. Invasive arterial pressure signals were processed using a validated artefact detection algorithm ( 13 ). The preprocessing workflow, including time alignment, artifact filtering, and feature selection steps, is illustrated in Supplementary Fig. 1. After preprocessing, 21,510 synchronized intraoperative observations (approximately 623,000 data points) were available for analysis. 2.4 Unsupervised identification of interaction states To identify deviations from the prevailing multivariate physiological structure, an unsupervised analytical framework was applied to the synchronized dataset. This approach evaluates intraoperative physiology as an interacting system rather than as isolated variables. Observations deviating from the dominant multivariate pattern were quantified using an anomaly score generated by an Extended Isolation Forest model. Mathematical details and parameter configurations are provided in Supplementary Fig. 1. 2.5 Definition of cardiopulmonary interaction states K-means clustering (k = 2) was applied to anomaly scores in the training dataset to differentiate typical from atypical multivariate patterns. This approach separated observations into dominant (coupled) and atypical interaction patterns. Atypical observations were subsequently stratified according to concurrently measured cardiac output to support physiological interpretation, resulting in three emergent interaction states: Coupled physiological state Low-flow, high-resistance interaction state Hyperdynamic, low-resistance interaction state Cardiac output was not used to train the unsupervised model but was applied post hoc solely to support physiological interpretation of emergent regimes. Supplementary Fig. 1 provides a conceptual illustration of the overall analytical workflow. 2.6 Post hoc Evaluation of Signal Contributions To determine which routinely available signals most significantly contributed to interaction-state discrimination, multiple feature sets were defined according to clinically distinct monitoring domains, including patient monitor variables, ventilatory parameters, anesthetic dosing, BIS, and advanced hemodynamic monitoring. For each feature set, a Random Forest classifier was trained on the training dataset and subsequently applied to the test dataset to assess the separability of interaction states as defined by the unsupervised framework. Performance was evaluated using precision–recall analysis, appropriate for imbalanced class distributions. 3. Results 3.1 Study cohort and interaction-state distribution Among 193 screened patients, 70 satisfied the inclusion criteria. Unsupervised analysis revealed three physiologically distinct interaction states across 21,510 synchronized observations (Table 1 ). The low-flow/high-resistance state was characterized by reduced stroke volume and elevated systemic vascular resistance, consistent with vasoconstrictive compensation. Conversely, the hyperdynamic/low-resistance state exhibited increased stroke volume, decreased systemic vascular resistance, elevated heart rate, and lower diastolic arterial pressure, indicating a flow-dominant circulation. Cardiac index (CI) values corresponded with state classification: CI was less than 2.2 L/min/m² in the low-flow and high-resistance state and greater than 4.0 L/min/m² in the hyperdynamic state. Because PEEP values were uniformly low, PEEP was analyzed categorically (< 4 vs ≥ 4 mbar). Table 1 Numeric data are presented as median (IQR—interquartile range). The dataset consists of 21,510 pairs of intraoperative monitoring data from 70 patients. Abbreviations: ART-SBP, ART-DBP – Systolic and diastolic invasive blood pressure; HR – Plethysmographic heart rate; BT – Body temperature; FiO₂, FeO₂ – Fraction of inspired and expired oxygen; RR – Respiratory rate; PEEP – Positive end-expiratory pressure; PIP – Peak inspiratory pressure; MAWP – Mean airway pressure; PPLAT – Plateau pressure; EtCO₂ – End-tidal CO₂ measured via infrared spectrometry capnography; PPF20-RATE – Propofol infusion rate (20 mg/mL); RFTN20-RATE – Remifentanil infusion rate (20 mcg/mL); EV1000.CO – Cardiac output (EV1000 device); EV1000.CI – Cardiac index (EV1000 device); SVR – Systemic vascular resistance; SV –Stroke volume; CVP – Central venous pressure (measured using the EV1000 device) Coupled physiological state Low-flow/high-resistance state Hyperdynamic/low-resistance state (N = 16,816) (N = 3,594) ( N = 1,100) Monitoring Parameters Median (IQR) Median (IQR) p-value 2 Median (IQR) p-value 2 Age, years 58 ( 16 ) 60 ( 15 ) < 0.001 46 ( 15 ) < 0.001 Sex < 0.001 < 0.001 Female 5,225 (31%) 3,119 (87%) 82 (7.5%) Male 11,591 (69%) 475 (13%) 1,018 (93%) Height, cm 164 ( 11 ) 155 ( 8 ) < 0.001 169 ( 6 ) < 0.001 Weight, kg 64 ( 17 ) 58 ( 12 ) < 0.001 73 ( 16 ) < 0.001 BT, °C 35.90 (0.80) 35.90 (0.70) 0.007 36.30 (0.60) < 0.001 Preop Hb, g/dl 13.00 (2.90) 13.00 (1.70) < 0.001 12.10 (5.20) < 0.001 Preop Alb, g/dL 4.00 (0.50) 4.00 (0.60) < 0.001 4.10 (0.50) 0.034 HR (/min) 69 (18) 64 ( 12 ) < 0.001 81 (21) < 0.001 ART-SBP, (mmHg) 117 (27) 125 (23) < 0.001 118 (38) 0.006 ART-DBP, (mmHg) 60 ( 13 ) 70 ( 16 ) < 0.001 57 ( 9 ) < 0.001 Minute Volume, (ml/kg/min) 87 (22) 85 (30) < 0.001 89 (24) < 0.001 Tidal Volume, (ml/kg) 5.99 (1.39) 5.88 (1.34) < 0.001 5.81 (0.83) < 0.001 RR, (/min) 14 ( 3 ) 14 ( 2 ) 0.001 17 ( 5 ) < 0.001 ETCO2, (mmHg) 35 ( 3 ) 34 ( 3 ) < 0.001 35 ( 2 ) 0.7 FIO2, (%) 36 ( 2 ) 36 ( 3 ) < 0.001 34 ( 2 ) < 0.001 FEO2, (%) 30 ( 2 ) 31 ( 2 ) < 0.001 29 ( 3 ) < 0.001 SPO2, (%) 100 (0) 100 (0) < 0.001 100 (0) < 0.001 PEEP, (mbar) < 0.001 < 0.001 <4 mbar 12,952 (77%) 3,175 (88%) 735 (67%) ≥4 mbar 3,893 (23%) 419 (12%) 365 (33%) PIP, (mbar) 13 ( 5 ) 12 ( 8 ) 0.032 15 ( 5 ) < 0.001 PPLAT, (mbar) 12 ( 6 ) 11 ( 8 ) 0.9 13 ( 5 ) 0.001 MAWP, (mbar) 4 ( 3 ) 3 ( 2 ) < 0.001 4 ( 3 ) < 0.001 PPF20-RATE, (mg/kg/min) 0.101 (0.035) 0.095 (0.029) < 0.001 0.101 (0.052) 0.7 RFTN20-RATE, (mcg/kg/min) 0.15 (0.08) 0.17 (0.06) < 0.001 0.18 (0.09) < 0.001 BIS 42 ( 10 ) 40 ( 8 ) < 0.001 44 ( 12 ) < 0.001 CVP, (cmH2O) 6 ( 4 ) 6 ( 2 ) < 0.001 8 ( 4 ) < 0.001 SV, (mL/beat) 76 (31) 49 ( 8 ) < 0.001 109 (41) < 0.001 SVR, (dn-s/cm5) 1,103 (453) 2,085 (401) < 0.001 656 (93) < 0.001 EV1000.CO, (L/min) 5.20 (2.10) 3.10 (0.60) < 0.001 8.60 (1.70) < 0.001 EV1000.CI, (L/min/m2) 3.11 (1.16) 2.08 (0.44) < 0.001 4.88 (0.75) < 0.001 Wilcoxon rank sum test 3.2 Separability using routine monitoring signals To assess whether interaction states could be distinguished using subsets of routinely available intraoperative variables, Random Forest classifiers were applied post hoc (Table 2 ). Models utilizing only mechanical ventilation parameters achieved strong discrimination for the low-flow/high-resistance state (AUPRC 0.906) and moderate discrimination for the hyperdynamic/vasodilated state (AUPRC 0.739). Including anesthetic infusion rates improved the separability of the hyperdynamic state (AUPRC 0.822). Importantly, models excluding advanced hemodynamic variables (cardiac output and cardiac index) retained robust performance. These findings indicate that Interaction states can be identified from routine intraoperative signals without reliance on invasive cardiac output monitoring. Table 2 Performance of the Random Forest model in distinguishing between normal and anomalous intraoperative hemodynamic states based on monitoring parameters. PM includes demographic data, basic monitoring parameters, preoperative hemoglobin, and albumin levels. Additionally, it incorporates mechanical ventilation (MV) parameters, propofol and remifentanil infusion parameters (P & R), Bispectral Index (BIS), and hemodynamic monitoring using the EV1000 device. Model Low Cardiac Output Anomaly High Cardiac Output Anomaly Precision Recall F1 Score AUROC PR Precision Recall F1 Score AUROC PR PM 0.985 0.99 0.987 0.879 0.987 0.995 0.991 0.711 PM + Hb + Alb 0.986 0.99 0.988 0.887 0.985 0.995 0.99 0.69 PM + Hb + Alb + MV 0.989 0.991 0.99 0.906 0.987 0.996 0.992 0.739 PM + Hb + Alb + MV + P + R 0.987 0.992 0.989 0.898 0.993 0.996 0.994 0.822 PM + Hb + Alb + MV + P + R + BIS 0.988 0.989 0.989 0.891 0.989 0.998 0.994 0.804 PM + Hb + Alb + MV + P + R + BIS + EV1000 0.997 0.995 0.996 0.959 0.996 0.999 0.997 0.927 3.3 Feature Contributions and Physiological Signatures Feature importance analysis (Figs. 1 and 2 ) identified distinct physiological signatures associated with each interaction state. In the low-flow and high-resistance state, diastolic arterial pressure was a significant contributor. Systemic vascular resistance also played a major role, while stroke volume provided additional explanatory value. Anthropometric and baseline variables, including weight, height, and preoperative hemoglobin, demonstrated measurable contributions. Ventilatory parameters such as peak inspiratory pressure and tidal volume contributed to state discrimination, although their influence was secondary to resistance-related and patient-specific factors. In the hyperdynamic and vasodilated state, heart rate and arterial pressure variables were the dominant contributors, consistent with a flow-dominant and low-resistance circulation. Body temperature and hemoglobin also made relevant contributions. Collectively, these findings suggest that interaction states reflect both ventilatory-hemodynamic coupling and patient-specific structural and hematologic characteristics. 4. Discussion This study demonstrates that intraoperative cardiopulmonary instability during general anesthesia can be conceptualized as a disruption of heart–lung coupling rather than as abnormalities in isolated physiological variables. The application of the Extended Isolation Forest (EIF) algorithm to synchronized routine monitoring data identified two reproducible interaction states: low-flow/high-resistance and hyperdynamic/vasodilated. These states were physiologically coherent and distinct from normal coupled physiology. Notably, detection was achievable using standard intraoperative signals, such as arterial pressure, ventilatory parameters, and anesthetic dosing, without reliance on advanced cardiac output monitoring. 4.1 Multivariate Physiological Nature of Cardiopulmonary Interaction Anomalies The EIF method identified deviations from the dominant multivariate structure of coupled cardiopulmonary physiology without relying on predefined instability thresholds. This approach posits that instability results from altered interactions among physiological subsystems rather than from excursions in individual variables. Correlation analysis stratified by interaction state revealed systematic physiological reorganization (Fig. 3 ). In the low-flow and high-resistance state, cardiac output was inversely correlated with systemic vascular resistance (r ≈ − 0.41), which is consistent with resistance-driven compensation rather than primary pump failure. Propofol–systemic vascular resistance (SVR) coupling was minimal in low-resistance states (r ≈ − 0.05) but pronounced in high-resistance states (r ≈ − 0.56), indicating increased susceptibility to anesthetic-induced hypotension when vascular tone is predominant. Central venous pressure was positively correlated with stroke volume in the high-resistance state (r ≈ + 0.57), but not in the hyperdynamic state (r ≈ − 0.06). This pattern indicates preserved preload responsiveness in resistance-dominant physiology and functional preload decoupling in flow-dominant conditions. A state-dependent inversion in the relationship between hemoglobin and cardiac output was observed. Hemoglobin was positively correlated with cardiac output in low-flow states (r ≈ + 0.57), consistent with compensation for oxygen delivery. In hyperdynamic states, the correlation was negative (r ≈ − 0.53), suggesting that viscosity-related mechanisms modulate flow. These findings support the interpretation that hemoglobin acts as a state-dependent circulatory modulator rather than a fixed determinant of flow. 4.2 State-Dependent Ventilatory Effects and Physiological Tipping Behavior A central finding of this study is that the relationship between ventilatory load and hemodynamic output is state-dependent and nonlinear. Figure 4 demonstrates distinct response geometries across interaction states, consistent with established heart-lung physiology. 4.2.1 Low-flow / high-resistance state: afterload modulation with saturation In the low-flow/high-resistance interaction state, moderate increases in peak inspiratory pressure (PIP), mean airway pressure (MAWP), and tidal volume were associated with modest increases in cardiac output (CO) and cardiac index (CI), followed by a plateau. This pattern is consistent with partial ventricular unloading during positive-pressure ventilation, reflecting a combined reduction in effective left ventricular afterload and preload modulation under conditions of elevated systemic resistance ( 15 ). The subsequent plateau suggests a saturation of this compensatory effect, occurring as preload limitation begins to counterbalance the benefits of afterload reduction. 4.2.2 Coupled physiology: preserved buffering capacity In the coupled physiology state, variation in ventilatory parameters resulted in minimal changes in CO, CI, and SV across the observed ranges. This flat response profile is consistent with preserved cardiopulmonary coupling and effective buffering, such that moderate ventilatory perturbations do not substantially alter forward flow. 4.2.3 Hyperdynamic / vasodilated state: preload sensitivity and nonlinear decline In the hyperdynamic/vasodilated interaction state, ventilatory increases were associated with an initial mild rise in flow, followed by a marked decline in CO, CI, and SV beyond higher PIP and tidal volume ranges. This pattern is consistent with increased preload sensitivity: elevated intrathoracic pressure reduces venous return and right ventricular filling, leading to transition into a flow-limited regime ( 10 ). In this state, ventilatory load acts as a destabilizing factor rather than a supportive one. Together, these findings support a regime-dependent interpretation of intraoperative instability: identical ventilatory inputs may enhance, preserve, or impair forward flow depending on the underlying cardiopulmonary coupling state. 4.3 Interpretation of Feature Contributions In the low-flow/high-resistance state (Fig. 1 ), invasive diastolic arterial pressure is identified as a key feature. This finding suggests compensatory vasoconstriction that preserves perfusion pressure despite reduced forward flow( 16 ). In contrast, the hyperdynamic/low-resistance state (Fig. 2 ) represents a flow-dominant regime in which plethysmographic heart rate, systolic and diastolic arterial pressures, and body temperature were primary contributors. Elevated systolic pressure signifies enhanced stroke volume-arterial coupling, whereas lower diastolic pressure reflects accelerated peripheral runoff resulting from decreased systemic vascular resistance( 16 ). Increased body temperature corresponds to heightened metabolic demand and peripheral vasodilation, which are frequently observed in septic hyperdynamic conditions. 4.4 Clinical and Scientific Implications The interaction-centered framework reconceptualizes intraoperative instability as a state-dependent of cardiopulmonary coupling that can be detected using routine monitoring data. This approach facilitates differentiation among resistance-dominant low-flow states, preload-sensitive hyperdynamic states, and stable coupled physiology, reducing dependence on invasive cardiac output monitoring. Unsupervised regime detection is intended to augment, rather than replace, clinical judgment by providing may assist in earlier recognition of physiological pattern shifts and supporting more targeted ventilatory or vasoactive interventions. From a scientific perspective, this work integrates unsupervised anomaly detection with established cardiopulmonary physiology, thereby supporting the development of interpretable and physiology-aligned monitoring algorithms in anesthesia. 4.5 Study Limitations This retrospective, single-center analysis may limit generalizability. The absence of detailed time-stamped annotations for fluid administration, vasoactive interventions, and bleeding events constrained causal interpretation of regime transitions. Limited PEEP variability restricted the assessment of PEEP-dependent effects. Additionally, baseline cardiac function data (e.g., echocardiography, diastolic, and right ventricular function) were unavailable, limiting mechanistic stratification of preload sensitivity and ventricular reserve. 4.6. Future Directions Future work should prospectively validate the interaction-state structure across institutions, including higher-risk populations and broader ventilatory strategies. Incorporating detailed temporal annotations (fluid boluses, vasoactive adjustments, recruitment maneuvers, bleeding events) would enable higher-resolution mapping of transitions into and out of interaction-failure states. Extending the framework to time-aware models may quantify transition dynamics (e.g., the rate of deterioration in coupling under rising ventilatory load) and inspire confidence in advancing physiological interpretability. 5. Conclusion Unsupervised analysis of routine intraoperative signals demonstrates that anomalies in cardiopulmonary interaction constitute a multivariate physiological phenomenon. Distinct coupling regimes, such as low-output/high-resistance and hyperdynamic/low-resistance states, display state-dependent ventilatory effects and tipping behavior. These findings support a state-dependent interpretation of intraoperative physiology, in which distinct cardiopulmonary interaction patterns can be identified using routine monitoring data. Further prospective validation is required to determine clinical utility. Declarations Author Contributions (CRediT) Albion Dervishi: Conceptualization, Methodology, Formal Analysis, Investigation, Visualization, Writing – Original Draft, Writing – Review & Editing. Declaration of Competing Interest The author declares no competing interests. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Ethics Approval The data were obtained from the VitalDB database, originally collected under Institutional Review Board approval (No. H-1408-101-605) at Seoul National University Hospital. The use of fully de-identified publicly available data did not require additional ethical approval. Data Availability The data supporting this study are publicly available from the VitalDB database (https://physionet.org/content/vitaldb/1.0.0/)(12)(17) References Masud F, Gheewala G, Giesecke M, Suarez EE, Ratnani I. Cardiogenic Shock in Perioperative and Intraoperative Settings: A Team Approach. Methodist Debakey Cardiovasc J. 2020. 10.14797/mdcj-16-1-e1 . Belletti A, Lomivorotov VV, Zangrillo A, Pieri M. Low cardiac output syndrome after adult cardiac surgery. Signa Vitae. 2025. 10.22514/sv.2025.167 . GUYTON AC, LINDSEY AW. Venous return at various right atrial pressures and the normal venous return curve. Am J Physiol. 1957;189(3). 10.1152/ajplegacy.1957.189.3.609 . Jellinek H, Krafft P, Fitzgerald RD, Schwarz S, Pinsky MR. Right atrial pressure predicts hemodynamic response to apneic positive airway pressure. Crit Care Med. 2000;28(3). 10.1097/00003246-200003000-00012 . Gertler R, Joshi GP. Modern understanding of intraoperative mechanical ventilation in normal and diseased lungs. Adv Anesth. 2010. 10.1016/j.aan.2010.07.002 . Pinsky MR. Functional Hemodynamic Monitoring: Current Concepts in Critical Care. Curr Opin Crit Care. 2014;20(3). Van Den Berg PCM, Jansen JRC, Pinsky MR. Effect of positive pressure on venous return in volume-loaded cardiac surgical patients. J Appl Physiol. 2002;92(3). 10.1152/japplphysiol.00487.2001 . Seely AJE, MacKlem PT. Complex systems and the technology of variability analysis. Crit Care. 2004;8(6). 10.1186/cc2948 . Moorman JR, Rusin CE, Lee H, Guin LE, Clark MT, Delos JB et al. Predictive monitoring for early detection of subacute potentially catastrophic illnesses in critical care. In: Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS. 2011. 10.1109/IEMBS.2011.6091407 Pinsky MR. Cardiopulmonary interactions: Physiologic basis and clinical applications. In: Annals of the American Thoracic Society. 2018. 10.1513/AnnalsATS.201704-339FR Liu FT, Ting KM, Zhou ZH. Isolation-based anomaly detection. ACM Trans Knowl Discov Data. 2012;6(1). 10.1145/2133360.2133363 . Lee HC, Park Y, Yoon S, Bin, Yang SM, Park D, Jung CW. VitalDB, a high-fidelity multi-parameter vital signs database in surgical patients. Sci Data. 2022;9(1):279. 10.1038/s41597-022-01411-5 . Du CH, Glick D, Tung A. Error-checking intraoperative arterial line blood pressures. J Clin Monit Comput. 2019;33(3). 10.1007/s10877-018-0167-7 . Magder S. Right Atrial Pressure in the Critically Ill: How to Measure, What Is the Value, What Are the Limitations? Chest. 2017. 10.1016/j.chest.2016.10.026 Kuhn BT, Bradley LA, Dempsey TM, Puro AC, Adams JY. Management of mechanical ventilation in decompensated heart failure. J Cardiovasc Dev Disease MDPI. 2016. 10.3390/jcdd3040033 . Nichols WW, O’Rourke MF, Vlachopoulos C, Hoeks AP, Reneman RS. McDonald’s blood flow in arteries theoreticxperimental and clinical principles. McDonald’s Blood Flow in Arteries, Sixth Edition: Theoretical, Experimental and Clinical Principles. 2011. 10.1111/j.1540-8175.1991.tb01207.x Goldberger AL, Amaral LAN, Glass L, Hausdorff JM, Ivanov PC, Mark RG et al. Components of a New Research Resource for Complex Physiologic Signals. Circulation. 2000;101(23). Additional Declarations No competing interests reported. Supplementary Files SupplementaryFigure1.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviewers invited by journal 18 Mar, 2026 Editor invited by journal 26 Feb, 2026 Editor assigned by journal 24 Feb, 2026 Submission checks completed at journal 24 Feb, 2026 First submitted to journal 24 Feb, 2026 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-8961496","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":608355875,"identity":"3df08ce4-b632-4e38-9f6c-cd757e80b95c","order_by":0,"name":"Albion Dervishi","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA+UlEQVRIiWNgGAWjYDACdh6GAwwMzEAWYwMzg4ENiNF4AK8WZlQtaWAGQS0MEC1g8jCYgVcLfzPvwQMfd1jLmc9Ibv5cUHDebm37YaAtNTbRuLRIHOZLODjzTLqxzI3ENukZBreTt51JBGo5lpbbgEvPYR6Dw7xthxNn8BxsY+YBajE7ANTC2HAYpxZ5qJZ6oJbmzzwG55LNzj/Er8UAqiVBgr2xQZrH4ICd2Q0CthiC/dKWbjiDvbENqCU5wewG0JYEPH6RO957+MPHNmt5CWb2x595/tjZm51Pf/jgQ40Nbu+jg0SwygRilYOAPSmKR8EoGAWjYGQAANkEYgJX19xCAAAAAElFTkSuQmCC","orcid":"","institution":"Medius CLINIC NÜRTINGEN - Academic Teaching Hospital of the University of Tübingen","correspondingAuthor":true,"prefix":"","firstName":"Albion","middleName":"","lastName":"Dervishi","suffix":""}],"badges":[],"createdAt":"2026-02-24 23:08:53","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8961496/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8961496/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105081622,"identity":"f5d791cd-e1f8-4b61-a405-78d5fad2880c","added_by":"auto","created_at":"2026-03-20 17:58:22","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":169036,"visible":true,"origin":"","legend":"\u003cp\u003eRelative feature importance for identifying the low-flow/high-resistance cardiopulmonary interaction state. Each subplot corresponds to a different feature-set configuration (routine monitoring variables and selected additions). Bars indicate impurity-based Random Forest importance, highlighting which physiological signals most strongly separate the low-flow/high-resistance interaction state from coupled physiology.\u003c/p\u003e","description":"","filename":"image1.png","url":"https://assets-eu.researchsquare.com/files/rs-8961496/v1/96c24aefcac93003ef763956.png"},{"id":105081620,"identity":"5f4fdaa1-0f51-4b49-8655-c04f89e0e3d5","added_by":"auto","created_at":"2026-03-20 17:58:22","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":161041,"visible":true,"origin":"","legend":"\u003cp\u003eRelative feature importance for identifying the hyperdynamic/vasodilated cardiopulmonary interaction state. Each subplot corresponds to a different feature-set configuration. Bars indicate impurity-based Random Forest importance, showing which physiological signals most strongly separate the hyperdynamic/vasodilated interaction state from coupled physiology.\u003c/p\u003e","description":"","filename":"image2.png","url":"https://assets-eu.researchsquare.com/files/rs-8961496/v1/9532f845408aa9b5e931c1e7.png"},{"id":105081621,"identity":"b69cc2a6-751a-4b68-a182-8d2ef44ff8af","added_by":"auto","created_at":"2026-03-20 17:58:22","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":366450,"visible":true,"origin":"","legend":"\u003cp\u003eSide-by-side correlation matrices for Low Cardiac Output (left) and High Cardiac Output (right) anomaly states. Spearman correlations were computed across intraoperative physiological features within each anomaly subgroup identified using Extended Isolation Forest and K-means clustering based on cardiac output (CO) from the EV1000 device.\u003c/p\u003e","description":"","filename":"image3.png","url":"https://assets-eu.researchsquare.com/files/rs-8961496/v1/fceb156c94832e7592c54498.png"},{"id":105081624,"identity":"d3c93603-974b-459c-ad33-c76f990621a3","added_by":"auto","created_at":"2026-03-20 17:58:22","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":433853,"visible":true,"origin":"","legend":"\u003cp\u003eillustrates state-dependent nonlinear relationships between ventilatory parameters - peak inspiratory pressure (PIP), mean airway pressure (MAWP), and tidal volume (TV) and hemodynamic outputs, including cardiac output (CO), cardiac index (CI), and stroke volume (SV).\u003c/p\u003e","description":"","filename":"image4.png","url":"https://assets-eu.researchsquare.com/files/rs-8961496/v1/afad7b206a76ec421b5dac8a.png"},{"id":105563146,"identity":"ee882ff2-8bc9-4ab0-b7e9-29b8b08b5a65","added_by":"auto","created_at":"2026-03-27 12:46:08","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2159358,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8961496/v1/72aab801-5906-4234-888a-f6ca77b743f5.pdf"},{"id":105081623,"identity":"6730dd0a-b690-4139-8e20-be94eb4a5c6c","added_by":"auto","created_at":"2026-03-20 17:58:22","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":865018,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFigure1.docx","url":"https://assets-eu.researchsquare.com/files/rs-8961496/v1/6428ab48286c02504bcb1312.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Unsupervised Identification of Intraoperative Cardiopulmonary Interaction States During General Anesthesia","fulltext":[{"header":"1. Introduction","content":"\u003cdiv id=\"Sec2\" class=\"Section2\"\u003e \u003ch2\u003e1.1 Clinical significance of cardiopulmonary interaction failure\u003c/h2\u003e \u003cp\u003eIntraoperative hemodynamic instability is a frequent and clinically significant challenge during general anesthesia, particularly among patients with limited cardiovascular reserve (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). Unrecognized rapid development of low-cardiac-output states may result in tissue hypoperfusion, postoperative organ dysfunction, prolonged intensive care unit stay(\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e), and increased mortality.\u003c/p\u003e \u003cp\u003eEarly identification of preload-dependent states, such as hypovolemia or vasodilated conditions, and their associated cardiopulmonary interactions is critical for preventing circulatory collapse. This need may be particularly relevant during trauma, ongoing hemorrhage, or anesthetic induction, when physiological reserve is limited (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e1.2 Heart-Lung Interaction as a State Variable\u003c/h2\u003e \u003cp\u003eIdentical ventilatory input can increase cardiac output in high-resistance states by unloading the left ventricle. However, in preload-dependent or vasodilated states, the same input may reduce cardiac output by limiting venous return(\u003cspan additionalcitationids=\"CR4 CR5 CR6\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e1.3 Physiological meaning emerges from signal interactions\u003c/h2\u003e \u003cp\u003eIntraoperative signals, such as arterial pressure, oxygenation, end-tidal CO₂, ventilatory parameters, and anesthetic dosing, are continuously recorded but are often interpreted as isolated variables. Clinically meaningful deterioration may arise from altered relationships among these signals rather than from threshold violations of individual parameters(\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). This perspective supports a framework in which physiological normality and deviation are defined by patterns of interaction among signals, rather than by isolated measurements(\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e1.4 Unsupervised identification of interaction states\u003c/h2\u003e \u003cp\u003eThis study evaluated whether unsupervised analysis of multivariate intraoperative data can identify coherent cardiopulmonary interaction states.\u003c/p\u003e \u003cp\u003eAn unsupervised analytical framework was applied to detect deviations from the prevailing multivariate physiological structure, thereby revealing altered coupling between ventilatory load and hemodynamic response(\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eUnlike supervised models that predict predefined outcomes, this approach seeks to uncover intrinsic interaction regimes within routine monitoring data. The objective is not to replace clinical judgment but to provide a physiology-based method for characterizing cardiopulmonary interaction states grounded in established cardiovascular and pulmonary mechanisms.\u003c/p\u003e \u003c/div\u003e"},{"header":"2. Materials and Methods","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Study Design and Data Source\u003c/h2\u003e \u003cp\u003eThis retrospective analysis utilized de-identified intraoperative physiological data from the VitalDB open-access database, which contains synchronized recordings from patients who underwent non-cardiac surgery at Seoul National University Hospital between August 2016 and June 2017(\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). Data collection was conducted prospectively with institutional review board approval (IRB No. H-1408-101-605) and are publicly available (NCT02914444). The dataset comprises continuous waveform and numeric recordings from multiple perioperative monitoring systems across 6,388 surgical cases.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Patient selection\u003c/h2\u003e \u003cp\u003eAdult patients were included if continuous intraoperative recordings were available from patient monitors, anesthesia machines, bispectral index monitoring, infusion pumps, and advanced hemodynamic monitoring (EV1000). To reduce confounding from major hemorrhage, cases with estimated blood loss\u0026thinsp;\u0026ge;\u0026thinsp;500 mL were excluded. After applying these criteria, 70 patients were included in the final analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Data Acquisition and Preprocessing\u003c/h2\u003e \u003cp\u003eThe analysis included six static variables (age, sex, height, weight, preoperative hemoglobin, and albumin) and 23 intraoperative dynamic parameters collected from multiple monitoring systems. These parameters comprised invasive arterial pressure, heart rate, oxygen saturation, body temperature, ventilatory variables, anesthetic infusion rates, depth of anesthesia (BIS), and advanced hemodynamic variables such as cardiac output, cardiac index, stroke volume, systemic vascular resistance, and central venous pressure.\u003c/p\u003e \u003cp\u003eDue to varying sampling frequencies across devices, only time-synchronized observations were retained. Implausible negative values and artefactual readings were excluded from the dataset. Body temperature values below 33\u0026deg;C were classified as sensor errors and removed. Invasive arterial pressure signals were processed using a validated artefact detection algorithm (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe preprocessing workflow, including time alignment, artifact filtering, and feature selection steps, is illustrated in Supplementary Fig.\u0026nbsp;1.\u003c/p\u003e \u003cp\u003eAfter preprocessing, 21,510 synchronized intraoperative observations (approximately 623,000 data points) were available for analysis.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Unsupervised identification of interaction states\u003c/h2\u003e \u003cp\u003eTo identify deviations from the prevailing multivariate physiological structure, an unsupervised analytical framework was applied to the synchronized dataset. This approach evaluates intraoperative physiology as an interacting system rather than as isolated variables.\u003c/p\u003e \u003cp\u003eObservations deviating from the dominant multivariate pattern were quantified using an anomaly score generated by an Extended Isolation Forest model. Mathematical details and parameter configurations are provided in Supplementary Fig.\u0026nbsp;1.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Definition of cardiopulmonary interaction states\u003c/h2\u003e \u003cp\u003eK-means clustering (k\u0026thinsp;=\u0026thinsp;2) was applied to anomaly scores in the training dataset to differentiate typical from atypical multivariate patterns. This approach separated observations into dominant (coupled) and atypical interaction patterns.\u003c/p\u003e \u003cp\u003eAtypical observations were subsequently stratified according to concurrently measured cardiac output to support physiological interpretation, resulting in three emergent interaction states:\u003c/p\u003e \u003cp\u003e \u003cul\u003e \u003cli\u003e \u003cp\u003eCoupled physiological state\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eLow-flow, high-resistance interaction state\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eHyperdynamic, low-resistance interaction state\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e \u003c/p\u003e \u003cp\u003eCardiac output was not used to train the unsupervised model but was applied post hoc solely to support physiological interpretation of emergent regimes.\u003c/p\u003e \u003cp\u003eSupplementary Fig.\u0026nbsp;1 provides a conceptual illustration of the overall analytical workflow.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e2.6 Post hoc Evaluation of Signal Contributions\u003c/h2\u003e \u003cp\u003eTo determine which routinely available signals most significantly contributed to interaction-state discrimination, multiple feature sets were defined according to clinically distinct monitoring domains, including patient monitor variables, ventilatory parameters, anesthetic dosing, BIS, and advanced hemodynamic monitoring.\u003c/p\u003e \u003cp\u003eFor each feature set, a Random Forest classifier was trained on the training dataset and subsequently applied to the test dataset to assess the separability of interaction states as defined by the unsupervised framework. Performance was evaluated using precision\u0026ndash;recall analysis, appropriate for imbalanced class distributions.\u003c/p\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Study cohort and interaction-state distribution\u003c/h2\u003e \u003cp\u003eAmong 193 screened patients, 70 satisfied the inclusion criteria. Unsupervised analysis revealed three physiologically distinct interaction states across 21,510 synchronized observations (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe low-flow/high-resistance state was characterized by reduced stroke volume and elevated systemic vascular resistance, consistent with vasoconstrictive compensation. Conversely, the hyperdynamic/low-resistance state exhibited increased stroke volume, decreased systemic vascular resistance, elevated heart rate, and lower diastolic arterial pressure, indicating a flow-dominant circulation.\u003c/p\u003e \u003cp\u003eCardiac index (CI) values corresponded with state classification: CI was less than 2.2 L/min/m\u0026sup2; in the low-flow and high-resistance state and greater than 4.0 L/min/m\u0026sup2; in the hyperdynamic state.\u003c/p\u003e \u003cp\u003eBecause PEEP values were uniformly low, PEEP was analyzed categorically (\u0026lt;\u0026thinsp;4 vs\u0026thinsp;\u0026ge;\u0026thinsp;4 mbar).\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\u003eNumeric data are presented as median (IQR\u0026mdash;interquartile range). The dataset consists of 21,510 pairs of intraoperative monitoring data from 70 patients. Abbreviations: ART-SBP, ART-DBP \u0026ndash; Systolic and diastolic invasive blood pressure; HR \u0026ndash; Plethysmographic heart rate; BT \u0026ndash; Body temperature; FiO₂, FeO₂ \u0026ndash; Fraction of inspired and expired oxygen; RR \u0026ndash; Respiratory rate; PEEP \u0026ndash; Positive end-expiratory pressure; PIP \u0026ndash; Peak inspiratory pressure; MAWP \u0026ndash; Mean airway pressure; PPLAT \u0026ndash; Plateau pressure; EtCO₂ \u0026ndash; End-tidal CO₂ measured via infrared spectrometry capnography; PPF20-RATE \u0026ndash; Propofol infusion rate (20 mg/mL); RFTN20-RATE \u0026ndash; Remifentanil infusion rate (20 mcg/mL); EV1000.CO \u0026ndash; Cardiac output (EV1000 device); EV1000.CI \u0026ndash; Cardiac index (EV1000 device); SVR \u0026ndash; Systemic vascular resistance; SV \u0026ndash;Stroke volume; CVP \u0026ndash; Central venous pressure (measured using the EV1000 device)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\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=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCoupled physiological state\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e \u003cp\u003eLow-flow/high-resistance state\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e \u003cp\u003eHyperdynamic/low-resistance state\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(N\u0026thinsp;=\u0026thinsp;16,816)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(N\u0026thinsp;=\u0026thinsp;3,594)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003e(\u003c/b\u003eN\u0026thinsp;=\u0026thinsp;1,100)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eMonitoring Parameters\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eMedian (IQR)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003eMedian (IQR)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003ep-value\u003c/b\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003eMedian (IQR)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003ep-value\u003c/b\u003e\u003csup\u003e\u003cem\u003e2\u003c/em\u003e\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge, years\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e58 (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e60 (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e46 (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5,225 (31%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3,119 (87%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e82 (7.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e11,591 (69%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e475 (13%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1,018 (93%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHeight, cm\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e164 (\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e155 (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e169 (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWeight, kg\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e64 (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e58 (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e73 (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBT, \u0026deg;C\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e35.90 (0.80)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e35.90 (0.70)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.007\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e36.30 (0.60)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePreop Hb, g/dl\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e13.00 (2.90)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13.00 (1.70)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e12.10 (5.20)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePreop Alb, g/dL\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4.00 (0.50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.00 (0.60)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.10 (0.50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.034\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHR (/min)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e69 (18)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e64 (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e81 (21)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eART-SBP, (mmHg)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e117 (27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e125 (23)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e118 (38)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.006\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eART-DBP, (mmHg)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e60 (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e70 (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e57 (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMinute Volume, (ml/kg/min)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e87 (22)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e85 (30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e89 (24)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTidal Volume, (ml/kg)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.99 (1.39)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e5.88 (1.34)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.81 (0.83)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRR, (/min)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e14 (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e14 (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e17 (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eETCO2, (mmHg)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e35 (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e34 (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e35 (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFIO2, (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e36 (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e36 (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e34 (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFEO2, (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e30 (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e31 (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e29 (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSPO2, (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e100 (0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e100 (0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e100 (0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePEEP, (mbar)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026lt;4 mbar\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12,952 (77%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3,175 (88%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e735 (67%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u0026ge;4 mbar\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3,893 (23%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e419 (12%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e365 (33%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePIP, (mbar)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e13 (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e12 (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.032\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e15 (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePPLAT, (mbar)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e12 (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e11 (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e13 (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMAWP, (mbar)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4 (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3 (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4 (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePPF20-RATE, (mg/kg/min)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.101 (0.035)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.095 (0.029)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.101 (0.052)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.7\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRFTN20-RATE, (mcg/kg/min)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.15 (0.08)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.17 (0.06)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.18 (0.09)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBIS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e42 (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e40 (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e44 (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCVP, (cmH2O)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6 (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6 (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e8 (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSV, (mL/beat)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e76 (31)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e49 (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e109 (41)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSVR, (dn-s/cm5)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1,103 (453)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2,085 (401)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e656 (93)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEV1000.CO, (L/min)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.20 (2.10)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3.10 (0.60)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e8.60 (1.70)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEV1000.CI, (L/min/m2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3.11 (1.16)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.08 (0.44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e4.88 (0.75)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eWilcoxon rank sum test\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\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=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Separability using routine monitoring signals\u003c/h2\u003e \u003cp\u003eTo assess whether interaction states could be distinguished using subsets of routinely available intraoperative variables, Random Forest classifiers were applied post hoc (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eModels utilizing only mechanical ventilation parameters achieved strong discrimination for the low-flow/high-resistance state (AUPRC 0.906) and moderate discrimination for the hyperdynamic/vasodilated state (AUPRC 0.739). Including anesthetic infusion rates improved the separability of the hyperdynamic state (AUPRC 0.822).\u003c/p\u003e \u003cp\u003eImportantly, models excluding advanced hemodynamic variables (cardiac output and cardiac index) retained robust performance. These findings indicate that Interaction states can be identified from routine intraoperative signals without reliance on invasive cardiac output monitoring.\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\u003ePerformance of the Random Forest model in distinguishing between normal and anomalous intraoperative hemodynamic states based on monitoring parameters. PM includes demographic data, basic monitoring parameters, preoperative hemoglobin, and albumin levels. Additionally, it incorporates mechanical ventilation (MV) parameters, propofol and remifentanil infusion parameters (P \u0026amp; R), Bispectral Index (BIS), and hemodynamic monitoring using the EV1000 device.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"9\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eModel\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c5\" namest=\"c2\"\u003e \u003cp\u003eLow Cardiac Output Anomaly\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"4\" nameend=\"c9\" namest=\"c6\"\u003e \u003cp\u003eHigh Cardiac Output Anomaly\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePrecision\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eRecall\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eF1 Score\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eAUROC PR\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003ePrecision\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eRecall\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eF1 Score\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003eAUROC PR\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.985\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.987\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.879\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.987\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.995\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.991\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.711\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePM\u0026thinsp;+\u0026thinsp;Hb\u0026thinsp;+\u0026thinsp;Alb\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.986\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.988\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.887\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.985\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.995\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.69\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePM\u0026thinsp;+\u0026thinsp;Hb\u0026thinsp;+\u0026thinsp;Alb\u0026thinsp;+\u0026thinsp;MV\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.989\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.991\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.906\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.987\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.996\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.992\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.739\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePM\u0026thinsp;+\u0026thinsp;Hb\u0026thinsp;+\u0026thinsp;Alb\u0026thinsp;+\u0026thinsp;MV\u0026thinsp;+\u0026thinsp;P\u0026thinsp;+\u0026thinsp;R\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.987\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.992\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.989\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.898\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.993\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.996\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.994\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.822\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePM\u0026thinsp;+\u0026thinsp;Hb\u0026thinsp;+\u0026thinsp;Alb\u0026thinsp;+\u0026thinsp;MV\u0026thinsp;+\u0026thinsp;P\u0026thinsp;+\u0026thinsp;R\u0026thinsp;+\u0026thinsp;BIS\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.988\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.989\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.989\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.891\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.989\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.998\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.994\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.804\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePM\u0026thinsp;+\u0026thinsp;Hb\u0026thinsp;+\u0026thinsp;Alb\u0026thinsp;+\u0026thinsp;MV\u0026thinsp;+\u0026thinsp;P\u0026thinsp;+\u0026thinsp;R\u0026thinsp;+\u0026thinsp;BIS\u0026thinsp;+\u0026thinsp;EV1000\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.997\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.995\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.996\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.959\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.996\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.999\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e0.997\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e0.927\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=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Feature Contributions and Physiological Signatures\u003c/h2\u003e \u003cp\u003eFeature importance analysis (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e1\u003c/span\u003e and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003e) identified distinct physiological signatures associated with each interaction state.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIn the low-flow and high-resistance state, diastolic arterial pressure was a significant contributor. Systemic vascular resistance also played a major role, while stroke volume provided additional explanatory value. Anthropometric and baseline variables, including weight, height, and preoperative hemoglobin, demonstrated measurable contributions. Ventilatory parameters such as peak inspiratory pressure and tidal volume contributed to state discrimination, although their influence was secondary to resistance-related and patient-specific factors.\u003c/p\u003e \u003cp\u003eIn the hyperdynamic and vasodilated state, heart rate and arterial pressure variables were the dominant contributors, consistent with a flow-dominant and low-resistance circulation. Body temperature and hemoglobin also made relevant contributions. Collectively, these findings suggest that interaction states reflect both ventilatory-hemodynamic coupling and patient-specific structural and hematologic characteristics.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eThis study demonstrates that intraoperative cardiopulmonary instability during general anesthesia can be conceptualized as a disruption of heart\u0026ndash;lung coupling rather than as abnormalities in isolated physiological variables.\u003c/p\u003e \u003cp\u003eThe application of the Extended Isolation Forest (EIF) algorithm to synchronized routine monitoring data identified two reproducible interaction states: low-flow/high-resistance and hyperdynamic/vasodilated. These states were physiologically coherent and distinct from normal coupled physiology. Notably, detection was achievable using standard intraoperative signals, such as arterial pressure, ventilatory parameters, and anesthetic dosing, without reliance on advanced cardiac output monitoring.\u003c/p\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Multivariate Physiological Nature of Cardiopulmonary Interaction Anomalies\u003c/h2\u003e \u003cp\u003eThe EIF method identified deviations from the dominant multivariate structure of coupled cardiopulmonary physiology without relying on predefined instability thresholds. This approach posits that instability results from altered interactions among physiological subsystems rather than from excursions in individual variables.\u003c/p\u003e \u003cp\u003eCorrelation analysis stratified by interaction state revealed systematic physiological reorganization (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn the low-flow and high-resistance state, cardiac output was inversely correlated with systemic vascular resistance (r\u0026thinsp;\u0026asymp;\u0026thinsp;\u0026minus;\u0026thinsp;0.41), which is consistent with resistance-driven compensation rather than primary pump failure. Propofol\u0026ndash;systemic vascular resistance (SVR) coupling was minimal in low-resistance states (r\u0026thinsp;\u0026asymp;\u0026thinsp;\u0026minus;\u0026thinsp;0.05) but pronounced in high-resistance states (r\u0026thinsp;\u0026asymp;\u0026thinsp;\u0026minus;\u0026thinsp;0.56), indicating increased susceptibility to anesthetic-induced hypotension when vascular tone is predominant.\u003c/p\u003e \u003cp\u003eCentral venous pressure was positively correlated with stroke volume in the high-resistance state (r\u0026thinsp;\u0026asymp;\u0026thinsp;+\u0026thinsp;0.57), but not in the hyperdynamic state (r\u0026thinsp;\u0026asymp;\u0026thinsp;\u0026minus;\u0026thinsp;0.06). This pattern indicates preserved preload responsiveness in resistance-dominant physiology and functional preload decoupling in flow-dominant conditions.\u003c/p\u003e \u003cp\u003eA state-dependent inversion in the relationship between hemoglobin and cardiac output was observed. Hemoglobin was positively correlated with cardiac output in low-flow states (r\u0026thinsp;\u0026asymp;\u0026thinsp;+\u0026thinsp;0.57), consistent with compensation for oxygen delivery. In hyperdynamic states, the correlation was negative (r\u0026thinsp;\u0026asymp;\u0026thinsp;\u0026minus;\u0026thinsp;0.53), suggesting that viscosity-related mechanisms modulate flow. These findings support the interpretation that hemoglobin acts as a state-dependent circulatory modulator rather than a fixed determinant of flow.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003e4.2 State-Dependent Ventilatory Effects and Physiological Tipping Behavior\u003c/h2\u003e \u003cp\u003eA central finding of this study is that the relationship between ventilatory load and hemodynamic output is state-dependent and nonlinear. Figure\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e4\u003c/span\u003e demonstrates distinct response geometries across interaction states, consistent with established heart-lung physiology.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec20\" class=\"Section3\"\u003e \u003ch2\u003e4.2.1 Low-flow / high-resistance state: afterload modulation with saturation\u003c/h2\u003e \u003cp\u003eIn the low-flow/high-resistance interaction state, moderate increases in peak inspiratory pressure (PIP), mean airway pressure (MAWP), and tidal volume were associated with modest increases in cardiac output (CO) and cardiac index (CI), followed by a plateau.\u003c/p\u003e \u003cp\u003eThis pattern is consistent with partial ventricular unloading during positive-pressure ventilation, reflecting a combined reduction in effective left ventricular afterload and preload modulation under conditions of elevated systemic resistance (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e). The subsequent plateau suggests a saturation of this compensatory effect, occurring as preload limitation begins to counterbalance the benefits of afterload reduction.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section3\"\u003e \u003ch2\u003e4.2.2 Coupled physiology: preserved buffering capacity\u003c/h2\u003e \u003cp\u003eIn the coupled physiology state, variation in ventilatory parameters resulted in minimal changes in CO, CI, and SV across the observed ranges. This flat response profile is consistent with preserved cardiopulmonary coupling and effective buffering, such that moderate ventilatory perturbations do not substantially alter forward flow.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section3\"\u003e \u003ch2\u003e4.2.3 Hyperdynamic / vasodilated state: preload sensitivity and nonlinear decline\u003c/h2\u003e \u003cp\u003eIn the hyperdynamic/vasodilated interaction state, ventilatory increases were associated with an initial mild rise in flow, followed by a marked decline in CO, CI, and SV beyond higher PIP and tidal volume ranges. This pattern is consistent with increased preload sensitivity: elevated intrathoracic pressure reduces venous return and right ventricular filling, leading to transition into a flow-limited regime (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e). In this state, ventilatory load acts as a destabilizing factor rather than a supportive one.\u003c/p\u003e \u003cp\u003eTogether, these findings support a regime-dependent interpretation of intraoperative instability: identical ventilatory inputs may enhance, preserve, or impair forward flow depending on the underlying cardiopulmonary coupling state.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec23\" class=\"Section2\"\u003e \u003ch2\u003e4.3 Interpretation of Feature Contributions\u003c/h2\u003e \u003cp\u003eIn the low-flow/high-resistance state (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e1\u003c/span\u003e), invasive diastolic arterial pressure is identified as a key feature. This finding suggests compensatory vasoconstriction that preserves perfusion pressure despite reduced forward flow(\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn contrast, the hyperdynamic/low-resistance state (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e2\u003c/span\u003e) represents a flow-dominant regime in which plethysmographic heart rate, systolic and diastolic arterial pressures, and body temperature were primary contributors. Elevated systolic pressure signifies enhanced stroke volume-arterial coupling, whereas lower diastolic pressure reflects accelerated peripheral runoff resulting from decreased systemic vascular resistance(\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e). Increased body temperature corresponds to heightened metabolic demand and peripheral vasodilation, which are frequently observed in septic hyperdynamic conditions.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003e4.4 Clinical and Scientific Implications\u003c/h2\u003e \u003cp\u003eThe interaction-centered framework reconceptualizes intraoperative instability as a state-dependent of cardiopulmonary coupling that can be detected using routine monitoring data. This approach facilitates differentiation among resistance-dominant low-flow states, preload-sensitive hyperdynamic states, and stable coupled physiology, reducing dependence on invasive cardiac output monitoring.\u003c/p\u003e \u003cp\u003eUnsupervised regime detection is intended to augment, rather than replace, clinical judgment by providing may assist in earlier recognition of physiological pattern shifts and supporting more targeted ventilatory or vasoactive interventions.\u003c/p\u003e \u003cp\u003eFrom a scientific perspective, this work integrates unsupervised anomaly detection with established cardiopulmonary physiology, thereby supporting the development of interpretable and physiology-aligned monitoring algorithms in anesthesia.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec25\" class=\"Section2\"\u003e \u003ch2\u003e4.5 Study Limitations\u003c/h2\u003e \u003cp\u003eThis retrospective, single-center analysis may limit generalizability. The absence of detailed time-stamped annotations for fluid administration, vasoactive interventions, and bleeding events constrained causal interpretation of regime transitions. Limited PEEP variability restricted the assessment of PEEP-dependent effects. Additionally, baseline cardiac function data (e.g., echocardiography, diastolic, and right ventricular function) were unavailable, limiting mechanistic stratification of preload sensitivity and ventricular reserve.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec26\" class=\"Section2\"\u003e \u003ch2\u003e4.6. Future Directions\u003c/h2\u003e \u003cp\u003eFuture work should prospectively validate the interaction-state structure across institutions, including higher-risk populations and broader ventilatory strategies. Incorporating detailed temporal annotations (fluid boluses, vasoactive adjustments, recruitment maneuvers, bleeding events) would enable higher-resolution mapping of transitions into and out of interaction-failure states.\u003c/p\u003e \u003cp\u003eExtending the framework to time-aware models may quantify transition dynamics (e.g., the rate of deterioration in coupling under rising ventilatory load) and inspire confidence in advancing physiological interpretability.\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eUnsupervised analysis of routine intraoperative signals demonstrates that anomalies in cardiopulmonary interaction constitute a multivariate physiological phenomenon. Distinct coupling regimes, such as low-output/high-resistance and hyperdynamic/low-resistance states, display state-dependent ventilatory effects and tipping behavior. These findings support a state-dependent interpretation of intraoperative physiology, in which distinct cardiopulmonary interaction patterns can be identified using routine monitoring data. Further prospective validation is required to determine clinical utility.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthor Contributions (CRediT)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAlbion Dervishi: Conceptualization, Methodology, Formal Analysis, Investigation, Visualization, Writing \u0026ndash; Original Draft, Writing \u0026ndash; Review \u0026amp; Editing.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDeclaration of Competing Interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe author declares no competing interests.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics Approval\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data were obtained from the VitalDB database, originally collected under Institutional Review Board approval (No. H-1408-101-605) at Seoul National University Hospital. The use of fully de-identified publicly available data did not require additional ethical approval.\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data supporting this study are publicly available from the VitalDB database (https://physionet.org/content/vitaldb/1.0.0/)(12)(17)\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eMasud F, Gheewala G, Giesecke M, Suarez EE, Ratnani I. Cardiogenic Shock in Perioperative and Intraoperative Settings: A Team Approach. Methodist Debakey Cardiovasc J. 2020. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.14797/mdcj-16-1-e1\u003c/span\u003e\u003cspan address=\"10.14797/mdcj-16-1-e1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBelletti A, Lomivorotov VV, Zangrillo A, Pieri M. Low cardiac output syndrome after adult cardiac surgery. Signa Vitae. 2025. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.22514/sv.2025.167\u003c/span\u003e\u003cspan address=\"10.22514/sv.2025.167\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGUYTON AC, LINDSEY AW. Venous return at various right atrial pressures and the normal venous return curve. Am J Physiol. 1957;189(3). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1152/ajplegacy.1957.189.3.609\u003c/span\u003e\u003cspan address=\"10.1152/ajplegacy.1957.189.3.609\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJellinek H, Krafft P, Fitzgerald RD, Schwarz S, Pinsky MR. Right atrial pressure predicts hemodynamic response to apneic positive airway pressure. Crit Care Med. 2000;28(3). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1097/00003246-200003000-00012\u003c/span\u003e\u003cspan address=\"10.1097/00003246-200003000-00012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGertler R, Joshi GP. Modern understanding of intraoperative mechanical ventilation in normal and diseased lungs. Adv Anesth. 2010. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.aan.2010.07.002\u003c/span\u003e\u003cspan address=\"10.1016/j.aan.2010.07.002\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePinsky MR. Functional Hemodynamic Monitoring: Current Concepts in Critical Care. Curr Opin Crit Care. 2014;20(3).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Den Berg PCM, Jansen JRC, Pinsky MR. Effect of positive pressure on venous return in volume-loaded cardiac surgical patients. J Appl Physiol. 2002;92(3). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1152/japplphysiol.00487.2001\u003c/span\u003e\u003cspan address=\"10.1152/japplphysiol.00487.2001\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSeely AJE, MacKlem PT. Complex systems and the technology of variability analysis. Crit Care. 2004;8(6). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/cc2948\u003c/span\u003e\u003cspan address=\"10.1186/cc2948\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoorman JR, Rusin CE, Lee H, Guin LE, Clark MT, Delos JB et al. Predictive monitoring for early detection of subacute potentially catastrophic illnesses in critical care. In: Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS. 2011. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/IEMBS.2011.6091407\u003c/span\u003e\u003cspan address=\"10.1109/IEMBS.2011.6091407\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePinsky MR. Cardiopulmonary interactions: Physiologic basis and clinical applications. In: Annals of the American Thoracic Society. 2018. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1513/AnnalsATS.201704-339FR\u003c/span\u003e\u003cspan address=\"10.1513/AnnalsATS.201704-339FR\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu FT, Ting KM, Zhou ZH. Isolation-based anomaly detection. ACM Trans Knowl Discov Data. 2012;6(1). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1145/2133360.2133363\u003c/span\u003e\u003cspan address=\"10.1145/2133360.2133363\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee HC, Park Y, Yoon S, Bin, Yang SM, Park D, Jung CW. VitalDB, a high-fidelity multi-parameter vital signs database in surgical patients. Sci Data. 2022;9(1):279. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41597-022-01411-5\u003c/span\u003e\u003cspan address=\"10.1038/s41597-022-01411-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDu CH, Glick D, Tung A. Error-checking intraoperative arterial line blood pressures. J Clin Monit Comput. 2019;33(3). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s10877-018-0167-7\u003c/span\u003e\u003cspan address=\"10.1007/s10877-018-0167-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMagder S. Right Atrial Pressure in the Critically Ill: How to Measure, What Is the Value, What Are the Limitations? Chest. 2017. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.chest.2016.10.026\u003c/span\u003e\u003cspan address=\"10.1016/j.chest.2016.10.026\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKuhn BT, Bradley LA, Dempsey TM, Puro AC, Adams JY. Management of mechanical ventilation in decompensated heart failure. J Cardiovasc Dev Disease MDPI. 2016. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/jcdd3040033\u003c/span\u003e\u003cspan address=\"10.3390/jcdd3040033\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNichols WW, O\u0026rsquo;Rourke MF, Vlachopoulos C, Hoeks AP, Reneman RS. McDonald\u0026rsquo;s blood flow in arteries theoreticxperimental and clinical principles. McDonald\u0026rsquo;s Blood Flow in Arteries, Sixth Edition: Theoretical, Experimental and Clinical Principles. 2011. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/j.1540-8175.1991.tb01207.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1540-8175.1991.tb01207.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoldberger AL, Amaral LAN, Glass L, Hausdorff JM, Ivanov PC, Mark RG et al. Components of a New Research Resource for Complex Physiologic Signals. Circulation. 2000;101(23).\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"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":"bmc-anesthesiology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bane","sideBox":"Learn more about [BMC Anesthesiology](http://bmcanesthesiol.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bane","title":"BMC Anesthesiology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Cardiopulmonary interaction, Hemodynamic instability, General anesthesia, Mechanical ventilation, Intraoperative monitoring, Physiological modeling","lastPublishedDoi":"10.21203/rs.3.rs-8961496/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8961496/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eIntraoperative hemodynamic instability may result from disruption of heart-lung interaction rather than isolated abnormalities in individual variables. We evaluated whether unsupervised analysis of routine intraoperative monitoring data can identify physiologically coherent cardiopulmonary interaction states.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eIntraoperative data from 70 adults (21,510 synchronized observations) in the VitalDB database were analyzed. An unsupervised analytical framework was applied to multivariate physiological signals. Identified deviations were stratified by concurrent cardiac output to define physiological interaction regimes. Feature importance and correlation analyses were used to characterize physiological patterns.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eThree interaction regimes emerged: coupled physiology, low-flow/high-resistance, and hyperdynamic/low-resistance states. The low-flow state demonstrated reduced stroke volume with elevated systemic vascular resistance, consistent with vasoconstrictive compensation. The hyperdynamic state was characterized by increased stroke volume, reduced resistance, tachycardia, and lower diastolic pressure. Ventilatory-hemodynamic relationships were state dependent, revealing nonlinear tipping behavior under rising ventilatory load. Interaction states were identifiable using routine monitoring signals without reliance on invasive cardiac output measurement.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eCardiopulmonary interaction states represent multivariate, state-dependent physiological patterns detectable through unsupervised analysis of routine intraoperative data.\u003c/p\u003e","manuscriptTitle":"Unsupervised Identification of Intraoperative Cardiopulmonary Interaction States During General Anesthesia","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-20 17:58:17","doi":"10.21203/rs.3.rs-8961496/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewersInvited","content":"","date":"2026-03-18T09:03:16+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-02-26T14:03:53+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-02-25T04:11:57+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-02-25T04:10:00+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Anesthesiology","date":"2026-02-24T22:55:20+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-anesthesiology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bane","sideBox":"Learn more about [BMC Anesthesiology](http://bmcanesthesiol.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bane","title":"BMC Anesthesiology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"b9f903ea-ff4c-4bb7-b13e-1525336e85d2","owner":[],"postedDate":"March 20th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-03-20T17:58:17+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-20 17:58:17","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8961496","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8961496","identity":"rs-8961496","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}