Discrepancies in Gross Tumor Volume and Motion between 4D-CT and Magnetic Resonance Imaging for Stereotactic Body Radiotherapy: A Phantom and Patient Study

Discrepancies in Gross Tumor Volume and Motion between 4D-CT and Magnetic Resonance Imaging for Stereotactic Body Radiotherapy: A Phantom and Patient Study | 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 Discrepancies in Gross Tumor Volume and Motion between 4D-CT and Magnetic Resonance Imaging for Stereotactic Body Radiotherapy: A Phantom and Patient Study Jinhu Chen, Hongchao You, Liming Ma, Xingwei An, Yong Yin, Xue Tao, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9157737/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 Precise delineation of gross tumor volume (GTV) and accurate assessment of respiratory motion are critical for stereotactic body radiotherapy (SBRT) of lung and liver tumors. This study aimed to systematically compare CT- and MRI-based tumor volume delineation and motion assessment using a dynamic phantom and patient data, with phantom-based 4D-MRI and real-time cine-MRI included alongside standard 4D-CT. Methods A 4D dynamic thorax phantom and twenty patients (ten lung, ten liver) were imaged using 4D-CT and MR-Linac MRI. Patients underwent T2W-MRI and cine-MRI, while a simulated 4D-MRI dataset was generated for the phantom only. Gross tumor volumes (GTVs) were contoured by six physicians on different image sets. Volume metrics, Dice Similarity Coefficient (DSC), and Hausdorff Distance (HD95) were used to compare delineations. Tumor motion amplitude from 4D-CT and cine-MRI was correlated. Results For patients, the GTV derived from T2W-MRI was significantly larger than that from 4D-CT (p = 0.028), with inter-modality delineation variability of DSC = 0.72 ± 0.11 and HD95 = 2.9 ± 1.5 mm. In the phantom, 4D-MRI and 4D-CT GTVs showed no significant difference (p = 0.251). Tumor motion amplitudes measured by cine-MRI were strongly correlated with those measured by 4D-CT (r > 0.9), and no significant overall difference was observed between the paired measurements (p = 0.626). However, real-time cine-MRI occasionally revealed motion excursions beyond the 4D-CT-based internal target volume. Conclusions T2W-MRI provides superior soft-tissue contrast for GTV delineation, often revealing a larger, potentially more accurate tumor volume. While 4D-CT and phantom-based 4D-MRI provide comparable motion estimates, real-time cine-MRI on the MR-Linac is essential to capture the full extent of tumor motion and guide adaptive SBRT for optimal accuracy. Volume variation 4D-MRI cine-MRI Moving targets MR-Linac Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 1. Background Stereotactic body radiation therapy (SBRT) has emerged as a highly effective primary treatment for inoperable lung and liver cancers, achieving excellent local control rates [ 1 , 2 ] . The efficacy of SBRT, however, is critically dependent on two factors: the precise delineation of the gross tumor volume (GTV) and the accurate management of intrafractional tumor motion, which is predominantly induced by respiration [ 3 , 4 ] . To address motion management, four-dimensional computed tomography (4D-CT) has been established as the clinical standard. 4D-CT captures the tumor position across the respiratory cycle, allowing for the construction of an internal target volume (ITV) that encompasses the entire motion path [ 5 , 6 ] . While 4D-CT reduces motion artifacts and improves target coverage compared to 3D-CT, its relatively poor soft-tissue contrast can challenge accurate GTV delineation, particularly for tumors adjacent to mediastinal or hepatic structures [ 7 ] . Furthermore, the ITV approach can lead to the irradiation of substantial volumes of healthy tissue, and 4D-CT relies on an external surrogate for respiratory phase correlation, which may not always perfectly reflect internal anatomy motion [ 8 ] . Additionally, tumor motion exhibits significant interfraction variability, meaning the motion amplitude and pattern observed during simulation may differ from those during subsequent treatment fractions, further complicating precise targeting [ 9 , 10 ] . In contrast, magnetic resonance imaging (MRI) offers superior soft-tissue contrast without ionizing radiation, making it highly attractive for radiotherapy target definition [ 11 ] . The development of cine-MRI and four-dimensional MRI (4D-MRI) has extended these advantages to motion assessment. Cine-MRI provides real-time, two-dimensional visualization of tumor motion [ 12 ] , while 4D-MRI generates a time-resolved volumetric dataset, often using internal navigators, such as the position of the diaphragm or a self-navigating radial k-space acquisition, as a more direct surrogate for respiratory motion to bin images into specific respiratory phases [ 13 , 14 ] . Prior studies have demonstrated the potential of 4D-MRI for improved GTV delineation and patient-specific motion analysis [ 15 ] . Despite these advancements, a critical, unresolved question remains: how do GTVs and motion assessments derived from 4D-CT and MRI systematically compare, and what are the clinical implications of any discrepancies for SBRT planning? Existing literature has often evaluated these modalities in isolation, and a direct, simultaneous comparison—especially one that validates findings against a known ground truth using a dynamic phantom—is lacking. This gap is significant because discrepancies in GTV definition directly impact treatment volume and dose to organs at risk. Therefore, this study systematically evaluated 4D-CT and MRI for lung and liver tumor delineation and motion assessment. Using a 4D dynamic thorax phantom with known ground-truth volume and motion together with 20 patients, we aimed to: 1) quantify GTV differences between T2W-MRI and 4D-CT; 2) evaluate inter-modality delineation consistency and observer variation by the Dice similarity coefficient (DSC) and Hausdorff distance (HD95); and 3) assess the correlation of tumor motion amplitudes between 4D-CT and cine MRI. 2. Methods 2.1 Study workflow The overall workflow of the study is shown in Fig. 1 , delineating the parallel paths for phantom and patient investigations. 2.2 Phantom study 2.2.1 Phantom and Motion Patterns A 4D Dynamic Thorax Phantom (CIRS MRgRT 008Z; CIRS Inc., Norfolk, VA, USA) was used to simulate three-dimensional motion. Constructed from nonferromagnetic materials and driven by piezoelectric motors, the phantom was MR-safe. The organically shaped tumor was actuated by two independent piezoelectric motors to produce superior–inferior (SI) motion, while anterior–posterior (AP) and left–right (LR) motion arose from the angular projection of tumor rotation. Five motion patterns were designed (Additional file 1: Table S1 ) to mimic clinically relevant motion, including regular sinusoidal motion with different amplitudes (Sin-10A, Sin-25A), complex multidirectional motion (Cos4AP, Sawtooth), and cardiac-induced motion (heart). Amplitude, period, and trajectory were selected from published ranges of respiratory- and cardiac-induced motion in thoracic and abdominal tumors [ 16 ] . These patterns served as controlled ground truth for comparing volume rendering and motion capture between 4D-CT and MRI. 2.2.2 Phantom Imaging Protocol 4D-CT The CT imaging data from the motion phantom were acquired using the Siemens CT Big Bore (Siemens Healthineers, Erlangen, Germany) and saved in DICOM format. The scan parameters were determined based on visual assessment of the tumors in the CT images. 3D and 4D-MRI : A simulated 4D-MRI dataset was generated for the phantom by acquiring five static 3D T2W-MRI volumes at predefined positions across one motion cycle. Detailed acquisition parameters are provided in Additional file 1: Additional Methods. Cine-MRI : Sagittal cine-MRI was acquired for the phantom using the default thorax protocol. Detailed sequence parameters are provided in Additional file 1: Additional Methods. 2.3 Patient Study 2.3.1 Patient Cohort Patients were prospectively enrolled (n = 20; 10 lung and 10 liver) with histologically confirmed primary/metastatic tumors planned for SBRT, eligible for both 4D-CT and MR-Linac imaging, and age ≥ 18 years; all provided written informed consent. 4D-CT was routine simulation, and additional MR-Linac imaging (T2W-MRI and cine-MRI) was acquired under a research protocol prior to fraction 1 (Table 1 ). Table 1 Patient and Tumor Characteristics Characteristic All Patients (n = 20) Lung Cancer (n = 10) Liver Cancer (n = 10) Demographics Age, years Median (Range) [59.9 (44.2–75.3)] Median (Range) [62.4 (54.5–75.3)] Median (Range) [57.4 (44.2–66.7)] Gender Male (n) /Female (n) [13/7] Male (n) / Female (n) [6/4] Male (n) / Female (n) [7/3] Tumor Characteristics Pathology Primary / Metastatic [16/4] Adenocarcinoma (7), Squamous cell carcinoma (3) HCC (6), Metastasis (4) Clinical Stage (AJCC 8th) † I (4), II (10), III (4), IV (2) I (8), II (2) I (4), II (1), III (3), IV (2) GTV Volume on 4D-CT, cm³ Median (Range) [31.2 (2.4-142.6)] Median (Range) [9.7 (2.4–15.1)] Median (Range) [52.4 (5.5-142.6)] Motion Amplitude (SI), mm* Mean ± SD [18.5 ± 4.7] Mean ± SD [16.7 ± 4.2] Mean ± SD [20.2 ± 5.6] Abbreviations: HCC, hepatocellular carcinoma; SI, superior–inferior; SD, standard deviation. † Clinical stage was assigned according to the corresponding disease-specific staging system used at the participating institution. * Tumor motion amplitude was measured as peak-to-peak displacement in the superior–inferior direction on 4D-CT. 2.3.2 Patient Imaging Protocol 4D-CT Data Acquisition Patient 4D-CT and 3D-CT data were acquired on a Siemens SOMATOM Confidence CT Big Bore scanner (Siemens Healthineers, Erlangen, Germany) at 120 kVp and reconstructed into 10 respiratory phases with 1.5-mm slice thickness, 1.0-mm increment, and a B30f convolution kernel. Respiratory traces were recorded using the Real-time Position Management system (Varian Medical Systems, Palo Alto, CA, USA). 3D MRI : Patients were imaged on an MR-Linac (T2W-MRI sequence). 4D-MRI 4D-MRI was not acquired in the patient cohort due to technical and workflow constraints at the time of the study, including the lack of a clinically streamlined 4D-MRI sequence on the MR-Linac and the extended scan time which was prohibitive within the clinical workflow. Cine MRI : During patient positioning, sagittal cine-MRI was acquired on the Unity MR-Linac (Elekta AB, Stockholm, Sweden) using the integrated 1.5T wide-bore MRI system and an 8-channel body array coil. Key parameters were TR/TE = 3.6/1.5 ms and a temporal resolution of 0.5 s per frame; additional acquisition details are provided in Additional file 1: Additional Methods. 2.3.3 Contouring for Phantom and Patients Image data were imported into MIM Maestro Workstation (version 7.1.2, MIM Software Inc., Cleveland, OH, USA) for GTV contouring. Six physicians independently delineated GTVs on each image set using a standardized contouring guideline, including fixed zoom and window/level settings (− 1024/300 HU for 4D-CT and 3020/1510 for T2W-MRI). For patient studies, only the primary tumor was contoured, and the ITV was defined as the union of all phase-specific 4D-CT GTVs without an additional margin. In addition, two senior radiation oncologists independently assessed tumor boundary clarity on co-registered 4D-CT and T2W-MRI and reached a consensus for each case. 2.4 Data and Statistical Analysis 2.4.1 Volumetric and Shape Comparison Metrics Volume metrics, the Dice Similarity Coefficient (DSC) [ 17 ], and the 95th percentile Hausdorff Distance (HD95) were used to compare delineations. Dice Similarity Coefficient (DSC) The DSC quantifies the volumetric overlap between two contours, A and B. It is defined as DSC(A, B) = 2|A ∩ B| / (|A| + |B|), where |·| denotes volume. DSC values range from 0 (no overlap) to 1 (perfect overlap). Hausdorff Distance (HD95) The Hausdorff Distance measures the maximum distance between two contours. To reduce sensitivity to single outlier points, we employed the 95th percentile Hausdorff Distance (HD95) [ 18 ] , which represents the 95th percentile of all distances between points on contour A and the nearest point on contour B (and vice versa), rather than the absolute maximum. A smaller HD95 indicates better geometric agreement. Coefficient of Variation (CV) The CV was used to assess the relative volumetric variability of GTVs across respiratory phases. It was calculated as CV = (σ / µ) × 100%, where σ is the standard deviation and µ is the mean GTV volume for a given tumor across phases. 2.4.2 Motion Analysis Tumor motion amplitude (peak-to-peak displacement in the primary direction) was measured from both 4D-CT and cine-MRI datasets. The correlation between motion amplitudes derived from these two modalities was assessed for both phantom and patient data. 2.4.3 Statistical analysis Statistical analyses were performed using SPSS 26.0 and R 4.2.2. Normality was assessed using the Shapiro–Wilk test. Paired t-tests or Wilcoxon signed-rank tests were used for within-subject comparisons, and independent t-tests or Mann–Whitney U tests were used for between-group comparisons, as appropriate. Categorical variables were compared using the Chi-square test or Fisher’s exact test. A two-sided p-value < 0.05 was considered statistically significant. 3. Results 3.1 Phantom Results 3.1.1 Phantom Volumetric Analysis The phantom GTV volume results from 4D-CT are summarized in Table 2 . The average absolute volume of GTVs across the 10 phases was 31.16 cm³, with a narrow 95% confidence interval (CI) ranging from 31.00 to 31.32 cm³. The average standard deviation (SD) of GTVs across the five simulated motion modes was 0.3176 cm³, with individual mode SDs ranging from 0.053 to 0.58 cm³. The mean coefficient of variation for the phantom GTVs was 1.79%, with a range of 0.85% to 5.76%, demonstrating high volumetric consistency in the controlled phantom setup. Table 2 Phantom GTV Volumes Across Motion Patterns and Imaging Modalities A. Absolute GTV Volumes (cm³) by Motion Pattern Imaging Modality Sin-10A Sin-25A Cos4AP Heart Sawtooth Mean ± SD 4D-CT 52.41 66.58 54.39 45.28 51.42 54.02 ± 7.86 3D-CT 51.88 65.61 52.87 47.15 50.53 53.61 ± 7.55 T2W-MRI 56.69 63.01 66.77 48.2 56.46 58.23 ± 7.51 4D-MRI 57.07 72.35 55.5 46.54 55.5 57.39 ± 9.36 B. Volumetric Variability Metrics for 4D-CT Across 10 Respiratory Phases Motion Pattern Mean GTV (cm³) SD (cm³) CV (%) 95% CI Lower 95% CI Upper Sin-10A 31.12 0.12 0.39 31.05 31.19 Sin-25A 31.25 0.18 0.58 31.15 31.35 Cos4AP 31.08 0.08 0.26 31.03 31.13 Heart 31.05 0.26 0.85 30.9 31.2 Sawtooth 31.1 0.05 0.17 31.07 31.13 Abbreviations: GTV, gross tumor volume; SD, standard deviation; CV, coefficient of variation; CI, confidence interval; CT, computed tomography; MRI, magnetic resonance imaging. Part A summarizes absolute GTV measurements across motion patterns and imaging modalities. Part B summarizes volumetric variability metrics derived from 4D-CT across 10 respiratory phases. In the motion phantom, the average GTV size from T2W-MRI across the five motion modes was 56.69 cm³ (range: 49.41–67.04 cm³). Similarly, the average GTV size from 4D-MRI was 57.39 cm³, with a slightly wider range of 45.79 to 68.99 cm³. The motion phantom was assessed using 4D-CT, 4D-MRI, and T2W-MRI. The results showed no significant difference (p = 0.2506) between the average GTV obtained from 4D-MRI imaging and the results from 4D-CT. For the phantom, the median GTV sizes across the four imaging modalities were as follows: 3D-CT GTV (median: 51.88 cm³, range: 47.15–65.61 cm³), T2W-MRI GTV (median: 56.69 cm³, range: 48.20-66.77 cm³), 4D-CT GTV (median: 52.40 cm³, range: 44.32–63.71 cm³), and 4D-MRI GTV (median: 55.50 cm³, range: 46.54–72.35 cm³). 3.1.2 Phantom Motion Analysis Based on the phantom results, 4D-CT may slightly underestimate the full motion envelope relative to MR-based techniques, as summarized in Fig. 5 B and Additional file 1: Table S2. 3.2 Patient Results 3.2.1 Patient Volumetric and Shape Analysis GTV Volume and Variability on 4D-CT: The detailed results for GTV volume are presented in Fig. 2 . As illustrated in Fig. 2 A, the average standard deviation (SD) of GTVs from the 10 phases was 1.54 cm³ (range: 1.02–3.45 cm³) for lung patients. For liver patients, the average SD was notably higher at 3.66 cm³, ranging from 2.58 to 5.68 cm³. Figure 2 B displays the coefficients of variation (CV) for GTV volumes across respiratory phases. Liver tumors demonstrated significantly greater volumetric variability during respiration than lung tumors (median CV: 8.95% vs 5.72%; Mann–Whitney U test, p = 0.0211). The volumetric variations of GTVs at different respiratory phases (GTV-0%, GTV-20%, GTV-50%, and GTV-70%) are detailed in Fig. 3 A. For liver patients, the mean GTV sizes were 50.63 cm³, 52.35 cm³, 55.78 cm³, and 53.76 cm³, respectively. For lung patients, the corresponding mean sizes were 7.54 cm³, 8.25 cm³, 9.27 cm³, and 8.15 cm³. Statistical analysis revealed that the volumetric variation among the 10 phases was insignificant (F = 0.005, P = 0.999). However, as shown in Fig. 3 B, the 3D-CT GTV size in lung patients (7.89 ± 4.99 cm³) was significantly smaller than the 4D-CT size (9.71 ± 6.04 cm³) (P = 0.001). A similar significant difference was observed in liver patients, where the 3D-CT GTV size (50.63 ± 61.90 cm³) was smaller than the 4D-CT size (69.83 ± 66.57 cm³) (P = 0.001). Comparison between MRI and CT-based GTVs: A comparative analysis of volumetric variation was performed on ten cases of lung and liver tumors preoperatively evaluated with 4D-CT, 3D-CT, and T2W-MRI imaging, as visualized in Fig. 5 A. A statistically significant difference (p = 0.0285) was observed between the average GTV obtained from MR imaging and the results from 4D-CT for both lung and liver patients. The inter-observer variability, reflected in the range of individual GTV volumes contoured by the six physicians, showed a pooled standard deviation of 1.8 cm³ for CT-based contours and 2.3 cm³ for MRI-based contours across the patient cohort. Qualitative assessment indicated that MR imaging demonstrated clearer and more precise tumor contours compared to contrast-enhanced CT. Furthermore, the data range illustrated in the figures clearly indicates that the volumetric range captured by 4D-CT is more extensive than that captured by T2W-MRI. The size of the GTV is significantly influenced by motion and delineation uncertainties, particularly for small mobile tumors. The relative difference in GTV, calculated as [(GTV₃DMR - GTV₃DCT) / GTV₃DCT × 100%], ranged from 2.2% to 26.2%, in absolute terms, this corresponded to volume differences ranging from 0.8 cm³ to 18.5 cm³, with the underlying 3D-CT GTVs ranging from 2.4 to 142.6 cm³, as detailed in Fig. 5 A. This represents a clinically significant variation in target volume for radiotherapy treatment planning. Table 3 further presents the HD95 and DICE values quantifying the inter-modality shape agreement for GTVs delineated by six physicians between 3D-CT and T2W-MRI. The average HD95 was 2.9 ± 1.5 mm, and the mean Dice Similarity Coefficient (DSC) was 0.72 ± 0.11. Table 3 Inter-modality Agreement Between 3D-CT and T2W-MRI Sin-10A Sin-25A Cos4AP Heart Sawtooth Liver Lung DSC 0.73 ± 0.11 0.63 ± 0.18 0.77 ± 0.16 0.61 ± 0.17 0.75 ± 0.11 0.59 ± 0.21 0.73 ± 0.15 (mean ± std) HD95(mm) 2.26 ± 1.2 4.26 ± 1.8 2.12 ± 1.1 5.54 ± 1.5 1.98 ± 1.3 6.35 ± 2.3 2.34 ± 1.2 (mean ± std) Abbreviations: DSC, Dice similarity coefficient; HD95, 95th percentile Hausdorff distance; CT, computed tomography; MRI, magnetic resonance imaging; SD, standard deviation. Values are presented as mean ± SD. Phantom columns summarize inter-modality agreement across predefined motion patterns, whereas Liver and Lung summarize patient-based comparisons between 3D-CT and T2-weighted MRI. 3.2.2 Patient Motion Analysis Tumor motion amplitudes measured by cine-MRI were strongly correlated with those measured by 4D-CT, as presented in Fig. 5 B and Additional file 1: Table S2. Nevertheless, real-time cine-MRI occasionally revealed motion excursions beyond the 4D-CT-based internal target volume. 3.3 Representative Image Comparison Figure 4 depicts a representative example of the difference in image quality between 4D-CT and 4D-MRI of the phantoms. This qualitative disparity can directly impact the accuracy of tumor delineation, especially for smaller tumors with substantial motion. The observed variation in GTV within the breathing cycle may be attributed to artifacts inherent in 4D image quality and intra-observer variation during delineation. 4. Discussion In radiation physics, dynamic phantom studies are widely employed to verify target volume and motion variations across different imaging modalities, utilizing predefined volumes, motion patterns, and amplitudes for assessment and comparison [ 19 – 22 ] . While phantoms simulate motion under idealized conditions, it is crucial to recognize that their results may diverge from those observed in actual patient studies. Internationally, several studies have focused on the development of dynamic MRI phantoms [ 23 , 24 ] . In this study, we conducted a comprehensive investigation using a dynamic thorax phantom compatible with both MRI and CT, complemented by clinical studies on patients with lung and liver tumors. Our primary objective was to collect clinical data on tumor motion and volume variation using 4D-CT, 4D-MRI, and real-time cine MRI. It is important to clarify that 4D MRI was not acquired in the patient cohort due to technical and workflow constraints at the time of the study (e.g., lack of a clinically streamlined sequence and prohibitive scan length). Therefore, our conclusions regarding 4D-MRI are derived from the phantom experiment, while the patient study focuses on comparing the primary MRI sequence used for online delineation (T2W-MRI) with the standard simulation modality (4D-CT). An observed phenomenon in our study was the systematically larger GTVs delineated on T2W-MRI compared to 4D-CT. This discrepancy can be partly understood by considering how motion impacts these modalities differently. A free-breathing T2W-MRI, with its longer acquisition time, produces a time-averaged intensity image of the tumor across many breathing cycles. The contoured GTV thus represents an intensity-averaged volume that inherently incorporates information about the tumor's shape over its motion path. In contrast, a GTV on a single-phase 4D-CT is a spatial snapshot at one moment, and even the ITV is a spatial union of such snapshots, not an intensity average. This fundamental difference in acquisition physics contributes to the systematic volume differences observed. A potential contributing factor to this discrepancy may be peritumoral edema, a well-documented imaging characteristic that is often more conspicuous on T2-weighted sequences due to their inherent sensitivity to fluid content, compared to CT [ 25 – 27 ] . This established principle offers a plausible explanation for this particular volumetric discrepancy, though it requires further direct histopathological correlation for confirmation in our cohort. In the phantom study, the target was delineated using both CT and T2-weighted MRI, with both modalities clearly displaying the target boundary, thereby confirming the reliability of the phantom model. However, in patient studies, a significant source of uncertainty in Gross Tumor Volume (GTV) delineation arises from the variable visualization of tumor boundaries, particularly between central and peripheral lesions [ 28 ] Peripheral lesions often have well-defined boundaries, whereas central lesions may be attached to normal hilar structures, making their margins indistinct. While CT often struggles to delineate the boundaries of central lung cancers, the superior soft tissue contrast of T2-weighted MRI provides clearer definition, underscoring its clinical value. We further evaluated inter-observer variation using the Dice similarity coefficient (DSC) and 95th percentile Hausdorff distance (HD95) among six physicians contouring GTVs on MRI and CT across different respiratory patterns. A clear discrepancy in target volume between MRI and CT was observed in the phantom, particularly under motion patterns with larger amplitudes. A similar trend was also seen in patients, especially in liver tumors with greater respiratory motion. These findings indicate that apparent tumor boundaries are highly motion-dependent, with more pronounced MRI–CT differences in liver tumors. This is consistent with Case et al., who reported smaller motion amplitudes on 4D-MRI than on 4D-CT [ 29 , 30 ] . Therefore, multimodal imaging integration is essential for more accurate target delineation. Breathing irregularities can introduce variations in tumor displacement throughout the breathing cycle during both 4D-CT and 4D-MRI acquisitions. For 4D-CT, the tumor motion is captured over a few bed positions or helical pitches, effectively representing an average over several breathing cycles [ 31 ] . For 4D-MRI, the initial phase (typically around 10 seconds, covering 2–3 cycles) determines the breathing amplitude used for the entire scan [ 32 ] . Consequently, the Internal Target Volume (ITV) derived from either 4D technique may not fully capture the tumor motion over extended periods, such as during a radiotherapy fraction. Investigating the discrepancies in tumor motion between 4D-CT and 4D-MRI, and evaluating the differences between simulated localization and real-time measurement, remain topics of active research [ 33 – 35 ] . This study used cine MRI on the MR-Linac to address this issue. Phantom and patient data showed that tumor motion amplitude measured on simulation 4D-CT better matched real-time motion during radiotherapy. Phantom results also confirmed that 4D-MRI-based simulated localization accurately represented tumor motion amplitude, supporting its clinical value. In some patients, cine MRI detected motion beyond the ITV from simulation 4D-CT, likely because the brief 4D scan failed to capture irregular breathing and complex motion. In such cases, adaptive MR-Linac replanning may be needed to maintain accuracy. For motion-managed radiotherapy, SBRT with real-time target monitoring is strongly recommended. A key finding of this study is the significant disparity in GTV volume variability between phantom and patient 4D-CT images. The coefficients of variation for lung and liver tumor GTVs in patients were 5.72% and 8.95%, respectively, substantially larger than the 1.79% variation observed in the phantom. These results indicate that real tumor motion and morphological changes during respiration are far more complex and must be carefully accounted for in clinical studies and practice. The decision to utilize a 1.5T MR-Linac in this study was informed by prior evidence suggesting that, for thoracic applications, 1.5T systems may offer advantages in mitigating respiratory motion artifacts and geometric distortion compared to 3.0T systems [ 36 ] . This study has several limitations. The primary limitation was the relatively small patient sample size, which may constrain the generalizability of the findings and the statistical power for subgroup analyses. All within-subject comparisons between imaging modalities were re-analyzed using appropriate paired statistical tests to ensure robust inference. A key technical limitation was the inability to acquire patient 4D-MRI due to workflow and sequence availability constraints at the time of the study, restricting our comparative 4D-MRI analysis to the controlled phantom environment. Finally, the patient cohort was selected based on suitability for both CT and MRI, which may introduce a selection bias by excluding tumors poorly visualized on either modality. 5. Conclusions T2W-MRI provides superior soft-tissue contrast for GTV delineation and may improve MRI-based target definition in MR-guided radiotherapy. While 4D-CT and phantom-based 4D-MRI provide reasonable motion estimates during simulation, they may not fully capture the entirety of real-time tumor motion during treatment. Real-time cine-MRI on the MR-Linac therefore provides important complementary information for adaptive SBRT. Integrating CT-based motion assessment with MRI-based soft-tissue visualization may further improve the precision of SBRT delivery. Abbreviations 4D-CT four-dimensional computed tomography 4D-MRI 4D magnetic resonance imaging AP anterior-posterior FSE Fast spin echo DSC Dice Similarity Coefficient HD95 Hausdorff distance ITV internal target volume MRI magnetic resonance imaging NSA Number of Signal Average OAR organs at risk SI superior-inferior SBRT Stereotactic body radiation therapy TR repetition time TE echo time Declarations Ethics approval and consent to participate This study was reviewed and approved by the Ethics Committee of Shandong First Medical University Affiliated Cancer Hospital (Approval No. SDTHEC-202409048) and was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants. Consent for publication Not applicable. Availability of data and materials The datasets generated and/or analyzed during the current study are not publicly available due to patient privacy and institutional restrictions but are available from the corresponding author on reasonable request and with permission from the participating institution(s). Competing interests The authors declare that they have no competing interests. Funding This work was supported in part by the National Natural Science Foundation of China (Grant No. 12475343), the Collaborative Academic Innovation Project of Shandong Cancer Hospital (TS004), and the Taishan Scholars Program of Shandong Province, China (Grant No. ts202408375). Authors' contributions JHC, HCY, and LM conceived and designed the study. JHC, HCY, LM, XA, XT, GL, and TH collected the data and performed image analysis and contouring. HCY, YY, XCL, and ZJL contributed to methodology development, data interpretation, and supervision. JHC drafted the manuscript. XCL and ZJL critically revised the manuscript for important intellectual content. All authors read and approved the final manuscript. Acknowledgements The authors gratefully acknowledge all patients who participated in this study and the clinical staff who supported image acquisition and data collection. Authors' information Not applicable. References Flentje M. Stereotactic Radiotherapy of Targets in the Lung and Liver. Strahlenther Onkol. 2001;177:645–55. Navarro-Martin A, Aso S, Cacicedo J, et al. Phase II Trial of SBRT for Stage I NSCLC: Survival, Local Control, and Lung Function at 36 Months. J Thorac Oncol. 2016;11:1101–11. Borm KJ, Oechsner M, Wiegandt M, Hofmeister A, Combs SE, Duma MN. Moving targets in 4D-CTs versus MIP and AIP: comparison of patients data to phantom data. BMC Cancer. 2018;18. Nakamura M, Narita Y, Matsuo Y, et al. Geometrical differences in target volumes between slow CT and 4D CT imaging in stereotactic body radiotherapy for lung tumors in the upper and middle lobe. Med Phys. 2008;35:4142–8. Rietzel E, Liu AK, Doppke KP, et al. Design of 4D treatment planning target volumes. Int J Radiat Oncol Biol Phys. 2006;66:287–95. Bradley JD, Nofal AN, Naqa IME, et al. Comparison of helical, maximum intensity projection (MIP), and averaged intensity (AI) 4D CT imaging for stereotactic body radiation therapy (SBRT) planning in lung cancer. Radiotherapy Oncol. 2006;81:264–8. Rebecca, Muirhead S, et al. Use of Maximum Intensity Projections (MIPs) for Target Outlining in 4D-CT Radiotherapy Planning. J Thorac Oncol. 2008;3:1433–8. Xue H, Artico J, Fontana M, Moon JC, Davies RH, Kellman P. Landmark Detection in Cardiac MRI by Using a Convolutional Neural Network. Radiol Artif Intell. 2021;3(5):e200197. Keall PJ, Mageras GS, Balter JM, Emery RS, Forster KM, Jiang SB, Kapatoes JM, Low DA, Murphy MJ, Murray BR, Ramsey CR, Van Herk MB, Vedam SS, Wong JW, Yorke E. The management of respiratory motion in radiation oncology report of AAPM Task Group 76. Med Phys. 2006;33(10):3874 – 900. 10.1118/1.2349696 . PMID: 17089851. Case RB, Sonke JJ, Moseley DJ, Kim J, Brock KK, Dawson LA. Inter- and intrafraction variability in liver position in non-breath-hold stereotactic body radiotherapy. Int J Radiat Oncol Biol Phys. 2009;75(1):302–8. 10.1016/j.ijrobp.2009.03.058 . Epub 2009 Jul 21. PMID: 19628342. Bourque AE, Bedwani S, Filion D, Carrier JF. A particle filter based autocontouring algorithm for lung tumor tracking using dynamic magnetic resonance imaging. Med Phys. 2016;43:5161–9. Ipsen S, Blanck O, Lowther NJ, et al. Towards real-time MRI-guided 3D localization of deforming targets for non-invasive cardiac radiosurgery. Phys Med Biol. 2016;61:7848. Bourque AE, Bedwani S, Carrier JF, et al. Particle Filter-Based Target Tracking Algorithm for Magnetic Resonance-Guided Respiratory Compensation: Robustness and Accuracy Assessment. Int J Radiat Oncol Biol Phys. 2018;100:325–34. Keall PJ, Mageras GS, Balter JM, Emery RS, Yorke E. The management of respiratory motion in radiation oncology report of AAPM Task Group 76. Med Phys. Med Phys. 2006;33:3874–900. Shimizu S, Shirato H, Ogura S, et al. Detection of lung tumor movement in real-time tumor-tracking radiotherapy. Int J Radiat Oncol Biol Phys. 2001;51:304–10. Spautz S, Haase L, Tschiche M, Makocki S, Richter C, Troost EGC, Stützer K. Comparison of 3D and 4D robustly optimized proton treatment plans for non-small cell lung cancer patients with tumour motion amplitudes larger than 5 mm. Phys Imaging Radiat Oncol. 2023;27:100465. PMID: 37449022; PMCID: PMC10338142. Zou KH, Warfield SK, Bharatha A, Tempany CM, Kaus MR, Haker SJ, Wells WM 3rd, Jolesz FA, Kikinis R. Statistical validation of image segmentation quality based on a spatial overlap index. Acad Radiol. 2004;11(2):178–89. 10.1016/s1076-6332(03)00671-8 . PMID: 14974593; PMCID: PMC1415224. Huttenlocher DP, Gregory A, Klanderman. Rucklidge. Comparing images using the Hausdorff distance. IEEE Trans Pattern Anal Mach Intell. 2002;15(9):850–63. Weidner A, Stengl C, Dinkel F, et al. An abdominal phantom with anthropomorphic organ motion and multimodal imaging contrast for MR-guided radiotherapy. Phys Med Biol. 2022;67:67. Niebuhr NI, Johnen W, Echner G, et al. The ADAM-pelvis phantom-an anthropomorphic, deformable and multimodal phantom for MRgRT. Phys Med Biol. 2019;64:04NT05. Nikolai J, Mickevicius X et al. Simultaneous motion monitoring and truth-in-delivery analysis imaging framework for MR-guided radiotherapy. Phys Med Biol. 2018. Bowen SR, Nyflot MJ, Herrmann C, et al. Imaging and dosimetric errors in 4D PET/CT-guided radiotherapy from patient-specific respiratory patterns: a dynamic motion phantom end-to-end study. Phys Med Biol. 2015;60:3731–46. Keizer DMDM, Kerkmeijer LGW, Willigenburg T, Lier ALHMWv, Boer JCJD. Prostate intrafraction motion during the preparation and delivery of MR-guided radiotherapy sessions on a 1.5T MR-Linac. Radiother Oncol 2020;151. Sergej, Schneider K et al. Commissioning of a 4D MRI phantom for use in MR-guided radiotherapy. Med Phys. 2018. Roy AE, Wells P. Volume definition in radiotherapy planning for lung cancer: how the radiologist can help. Cancer imaging: official publication Int Cancer Imaging Soc. 2006;6:116–23. Lampen-Sachar K, Zhao B, Zheng J, et al. Correlation between tumor measurement on Computed Tomography and resected specimen size in lung adenocarcinomas. Lung Cancer. 2012;75:332–5. Yu HM, Liu YF, Hou M, Liu J, Li XN, Yu JM. Evaluation of gross tumor size using CT, 18F-FDG PET, integrated 18F-FDG PET/CT and pathological analysis in non-small cell lung cancer. Eur J Radiol. 2009;72:104–13. Nestle U, De Ruysscher D, Ricardi U, et al. ESTRO ACROP guidelines for target volume definition in the treatment of locally advanced non-small cell lung cancer. Radiotherapy oncology: J Eur Soc Therapeutic Radiol Oncol. 2018;127:1–5. Case RB, Sonke JJ, Moseley DJ, Kim J, Brock KK, Dawson LA. Inter- and intrafraction variability in liver position in non-breath-hold stereotactic body radiotherapy. Int J Radiat Oncol Biol Phys. 2009;75:302–8. van de Lindt TN, Nowee ME, Janssen T, et al. Technical feasibility and clinical evaluation of 4D-MRI guided liver SBRT on the MR-Linac. Radiother Oncol. 2022;167:285–91. Stevens CW, Munden RF, Forster KM, et al. Respiratory-driven lung tumor motion is independent of tumor size, tumor location, and pulmonary function. Int J Radiat Oncol Biol Phys. 2001;51:62–8. Zhang J, Srivastava S, Wang C, et al. Clinical evaluation of 4D MRI in the delineation of gross and internal tumor volumes in comparison with 4D-CT. J Appl Clin Med Phys. 2019;20:51–60. Li G, Citrin D, Camphausen K, Mueller B, Burman C, Mychalczak B, Miller RW, Song Y. Advances in 4D medical imaging and 4D radiation therapy. Technol Cancer Res Treat. 2008;7(1):67–81. von Siebenthal M, Szekely G, Gamper U, Boesiger P, Lomax A, Cattin P. 4D MR imaging of respiratory organ motion and its variability. Phys Med Biol. 2007;52:1547–64. Slagowski JM, Ding Y, Aima M, et al. A modular phantom and software to characterize 3D geometric distortion in MRI. Phys Med Biol. 2020;65:195008. Liu X, Li Z, Rong Y, Cao M, Qiu J. A Comparison of the Distortion in the Same Field MRI and MR-Linac System With a 3D Printed Phantom. Front Oncol 2021;11. Additional Declarations No competing interests reported. Supplementary Files Additionalfile1.docx Additional file 1 File format: DOCX Title: Additional Methods and Tables Description: Detailed MRI acquisition parameters and supplementary Tables S1–S2 for the phantom and patient studies. Cite Share Download PDF Status: Under Review Version 1 posted Reviewers agreed at journal 28 Apr, 2026 Reviewers invited by journal 10 Apr, 2026 Editor assigned by journal 30 Mar, 2026 Submission checks completed at journal 27 Mar, 2026 First submitted to journal 26 Mar, 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-9157737","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":622012140,"identity":"af1abd0f-4839-469c-97b5-732d828b73a9","order_by":0,"name":"Jinhu Chen","email":"","orcid":"","institution":"Tianjin University","correspondingAuthor":false,"prefix":"","firstName":"Jinhu","middleName":"","lastName":"Chen","suffix":""},{"id":622012143,"identity":"9a19e5a0-a8e7-4200-845d-30dd73ef2d58","order_by":1,"name":"Hongchao You","email":"","orcid":"","institution":"Shandong First Medical University","correspondingAuthor":false,"prefix":"","firstName":"Hongchao","middleName":"","lastName":"You","suffix":""},{"id":622012146,"identity":"6e0123bc-7c82-4a00-a250-fc01e902a66a","order_by":2,"name":"Liming Ma","email":"","orcid":"","institution":"Shandong First Medical University","correspondingAuthor":false,"prefix":"","firstName":"Liming","middleName":"","lastName":"Ma","suffix":""},{"id":622012151,"identity":"06694a70-195e-4ae6-9962-fb5006151446","order_by":3,"name":"Xingwei An","email":"","orcid":"","institution":"Tianjin University","correspondingAuthor":false,"prefix":"","firstName":"Xingwei","middleName":"","lastName":"An","suffix":""},{"id":622012153,"identity":"aeee4e1d-8ee6-4745-a771-c41f40a78887","order_by":4,"name":"Yong Yin","email":"","orcid":"","institution":"Shandong Cancer Hospital and Institute, Shandong First Medical University","correspondingAuthor":false,"prefix":"","firstName":"Yong","middleName":"","lastName":"Yin","suffix":""},{"id":622012157,"identity":"7a111644-1c4b-47eb-8ddc-0163005c7c2c","order_by":5,"name":"Xue Tao","email":"","orcid":"","institution":"Cangzhou Hospital of Integrated Traditional Chinese and Western Medicine","correspondingAuthor":false,"prefix":"","firstName":"Xue","middleName":"","lastName":"Tao","suffix":""},{"id":622012162,"identity":"078261ee-06e5-4e8f-8106-fee304f3c1d8","order_by":6,"name":"Guangbo Liu","email":"","orcid":"","institution":"Cangzhou Hospital of Integrated Traditional Chinese and Western Medicine","correspondingAuthor":false,"prefix":"","firstName":"Guangbo","middleName":"","lastName":"Liu","suffix":""},{"id":622012169,"identity":"0e947df6-bbf3-4d58-a1bd-0fe693017409","order_by":7,"name":"Tingting Hu","email":"","orcid":"","institution":"Cangzhou Hospital of Integrated Traditional Chinese and Western Medicine","correspondingAuthor":false,"prefix":"","firstName":"Tingting","middleName":"","lastName":"Hu","suffix":""},{"id":622012170,"identity":"bb92447b-3162-4f99-9d4a-7319dfc6e444","order_by":8,"name":"Xuechun Liu","email":"","orcid":"","institution":"Yantai Yuhuangding Hospital, Qingdao University School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Xuechun","middleName":"","lastName":"Liu","suffix":""},{"id":622012176,"identity":"9dad32ec-c47a-4d04-8c49-c6eb8042d917","order_by":9,"name":"Zhenjiang Li","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAwUlEQVRIiWNgGAWjYBACPmYGBoMPDMwkaGEDKjacQZoWIGbmIU0LO/uDYps/1gz87WcPfvi4h0GeX+wAQYclGOe2pTNInMlLlpzxjMFw5uwEgloOGOc2HK7fwJBjxsxzgCHB4DZBLYwNxhZ/DjMY8L8xY/5DnBZmBmMGNqAWCaAtDMRpYWMw7AX55cYbY8meAxKE/cLPf/yZwQ9QiPXnGH74ccBGnl+agBaQRQZIHAmCykGA+QFRykbBKBgFo2DkAgB0WTYsUbFOUQAAAABJRU5ErkJggg==","orcid":"","institution":"Shandong Cancer Hospital and Institute, Shandong First Medical University","correspondingAuthor":true,"prefix":"","firstName":"Zhenjiang","middleName":"","lastName":"Li","suffix":""}],"badges":[],"createdAt":"2026-03-18 09:54:59","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9157737/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9157737/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107319708,"identity":"96a7aa2a-7ea5-4069-8711-790bd6523b23","added_by":"auto","created_at":"2026-04-20 10:20:20","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":676453,"visible":true,"origin":"","legend":"\u003cp\u003eSchematic overview of the study workflow. The phantom and patient cohorts followed parallel imaging pathways. The phantom was imaged under five distinct programmed motion patterns. Both cohorts underwent 3D-CT (free-breathing), 4D-CT, and 3D T2-weighted MRI (T2W-MRI). A simulated 4D-MRI dataset was generated for the phantom only. Cine-MRI was acquired for real-time motion assessment in both. †4D-MRI was not acquired in the patient cohort due to technical and workflow constraints.\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9157737/v1/97d340ede219c2a1d3328c2c.jpeg"},{"id":107319709,"identity":"cddc34ba-8bf5-4127-9427-e92f86581294","added_by":"auto","created_at":"2026-04-20 10:20:20","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":421786,"visible":true,"origin":"","legend":"\u003cp\u003eVolumetric variability of GTVs across 4D-CT respiratory phases in patients. (A) Standard deviation (SD) of GTV volumes for each patient. (B) Coefficient of variation (CV) of GTV volumes for lung and liver tumors. Liver tumors exhibited significantly greater volumetric variability than lung tumors (p = 0.0211).\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9157737/v1/fff1f473874f2b8dd55d8df5.jpeg"},{"id":107485427,"identity":"d2d81c00-a64c-4099-a0c2-95cc8a2abbb0","added_by":"auto","created_at":"2026-04-22 02:34:43","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":277062,"visible":true,"origin":"","legend":"\u003cp\u003eThe 4D-CT-defined target volume is larger than the 3D-CT volume, despite minimal respiratory-phase variation.(A) GTV volume across four respiratory phases shows no significant trend of change.(B) Direct comparison reveals that the volume encompassing motion (4D-CT) is significantly larger than the snapshot volume (3D-CT) for both lung and liver tumors (P = 0.001).\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-9157737/v1/df4704f8eeaee9f99538b680.png"},{"id":107319711,"identity":"cf7fc12c-0256-47ee-8640-129f0f4876da","added_by":"auto","created_at":"2026-04-20 10:20:20","extension":"jpeg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":872760,"visible":true,"origin":"","legend":"\u003cp\u003eRepresentative phantom images from CT- and MRI-based motion imaging. Representative axial and coronal views are shown for CT- and MRI-based acquisitions, with overlaid target contours. The figure illustrates the improved soft-tissue conspicuity of MRI relative to CT for target delineation in the motion phantom.\u003c/p\u003e","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-9157737/v1/88f3b4b4306dc52ec6ec62eb.jpeg"},{"id":107319713,"identity":"8e8840f7-a57d-42d9-b6f4-3d2aacb90dd3","added_by":"auto","created_at":"2026-04-20 10:20:20","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":487882,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of MRI- and CT-based target volumes and motion measurements. (A) Relative GTV differences between 3D MRI and 3DCT across patient and phantom cohorts. (B) Agreement of motion amplitude measurements between 4DCT and cine-MRI using correlation, Bland–Altman analysis, group-wise comparisons, and modality-specific summaries.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-9157737/v1/51b1ce6dcc817e426b1bb0e5.png"},{"id":107487840,"identity":"029a407e-d51e-4983-b5c7-f639d88b2f70","added_by":"auto","created_at":"2026-04-22 02:42:57","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3212805,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9157737/v1/35ad58ce-6873-4500-9d0c-e282bbe23947.pdf"},{"id":107319712,"identity":"76f3a375-66df-4daf-a604-4604e6e04158","added_by":"auto","created_at":"2026-04-20 10:20:20","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":28463,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAdditional file 1\u003c/strong\u003e\u003cbr\u003e\n\u003cstrong\u003eFile format:\u003c/strong\u003e DOCX\u003cbr\u003e\n\u003cstrong\u003eTitle:\u003c/strong\u003e Additional Methods and Tables\u003cbr\u003e\n\u003cstrong\u003eDescription:\u003c/strong\u003e Detailed MRI acquisition parameters and supplementary Tables S1–S2 for the phantom and patient studies.\u003c/p\u003e","description":"","filename":"Additionalfile1.docx","url":"https://assets-eu.researchsquare.com/files/rs-9157737/v1/83a3db9f55c4759894891f8f.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Discrepancies in Gross Tumor Volume and Motion between 4D-CT and Magnetic Resonance Imaging for Stereotactic Body Radiotherapy: A Phantom and Patient Study","fulltext":[{"header":"1. Background","content":"\u003cp\u003eStereotactic body radiation therapy (SBRT) has emerged as a highly effective primary treatment for inoperable lung and liver cancers, achieving excellent local control rates\u003csup\u003e[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]\u003c/sup\u003e. The efficacy of SBRT, however, is critically dependent on two factors: the precise delineation of the gross tumor volume (GTV) and the accurate management of intrafractional tumor motion, which is predominantly induced by respiration\u003csup\u003e[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eTo address motion management, four-dimensional computed tomography (4D-CT) has been established as the clinical standard. 4D-CT captures the tumor position across the respiratory cycle, allowing for the construction of an internal target volume (ITV) that encompasses the entire motion path\u003csup\u003e[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]\u003c/sup\u003e. While 4D-CT reduces motion artifacts and improves target coverage compared to 3D-CT, its relatively poor soft-tissue contrast can challenge accurate GTV delineation, particularly for tumors adjacent to mediastinal or hepatic structures\u003csup\u003e[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]\u003c/sup\u003e. Furthermore, the ITV approach can lead to the irradiation of substantial volumes of healthy tissue, and 4D-CT relies on an external surrogate for respiratory phase correlation, which may not always perfectly reflect internal anatomy motion\u003csup\u003e[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]\u003c/sup\u003e. Additionally, tumor motion exhibits significant interfraction variability, meaning the motion amplitude and pattern observed during simulation may differ from those during subsequent treatment fractions, further complicating precise targeting \u003csup\u003e[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn contrast, magnetic resonance imaging (MRI) offers superior soft-tissue contrast without ionizing radiation, making it highly attractive for radiotherapy target definition\u003csup\u003e[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]\u003c/sup\u003e. The development of cine-MRI and four-dimensional MRI (4D-MRI) has extended these advantages to motion assessment. Cine-MRI provides real-time, two-dimensional visualization of tumor motion\u003csup\u003e[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]\u003c/sup\u003e, while 4D-MRI generates a time-resolved volumetric dataset, often using internal navigators, such as the position of the diaphragm or a self-navigating radial k-space acquisition, as a more direct surrogate for respiratory motion to bin images into specific respiratory phases\u003csup\u003e[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]\u003c/sup\u003e. Prior studies have demonstrated the potential of 4D-MRI for improved GTV delineation and patient-specific motion analysis\u003csup\u003e[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eDespite these advancements, a critical, unresolved question remains: how do GTVs and motion assessments derived from 4D-CT and MRI systematically compare, and what are the clinical implications of any discrepancies for SBRT planning? Existing literature has often evaluated these modalities in isolation, and a direct, simultaneous comparison\u0026mdash;especially one that validates findings against a known ground truth using a dynamic phantom\u0026mdash;is lacking. This gap is significant because discrepancies in GTV definition directly impact treatment volume and dose to organs at risk.\u003c/p\u003e \u003cp\u003eTherefore, this study systematically evaluated 4D-CT and MRI for lung and liver tumor delineation and motion assessment. Using a 4D dynamic thorax phantom with known ground-truth volume and motion together with 20 patients, we aimed to: 1) quantify GTV differences between T2W-MRI and 4D-CT; 2) evaluate inter-modality delineation consistency and observer variation by the Dice similarity coefficient (DSC) and Hausdorff distance (HD95); and 3) assess the correlation of tumor motion amplitudes between 4D-CT and cine MRI.\u003c/p\u003e"},{"header":"2. Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Study workflow\u003c/h2\u003e \u003cp\u003eThe overall workflow of the study is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003e, delineating the parallel paths for phantom and patient investigations.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Phantom study\u003c/h2\u003e \u003cdiv id=\"Sec5\" class=\"Section3\"\u003e \u003ch2\u003e2.2.1 Phantom and Motion Patterns\u003c/h2\u003e \u003cp\u003eA 4D Dynamic Thorax Phantom (CIRS MRgRT 008Z; CIRS Inc., Norfolk, VA, USA) was used to simulate three-dimensional motion. Constructed from nonferromagnetic materials and driven by piezoelectric motors, the phantom was MR-safe. The organically shaped tumor was actuated by two independent piezoelectric motors to produce superior\u0026ndash;inferior (SI) motion, while anterior\u0026ndash;posterior (AP) and left\u0026ndash;right (LR) motion arose from the angular projection of tumor rotation.\u003c/p\u003e \u003cp\u003eFive motion patterns were designed (Additional file 1: Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e) to mimic clinically relevant motion, including regular sinusoidal motion with different amplitudes (Sin-10A, Sin-25A), complex multidirectional motion (Cos4AP, Sawtooth), and cardiac-induced motion (heart). Amplitude, period, and trajectory were selected from published ranges of respiratory- and cardiac-induced motion in thoracic and abdominal tumors \u003csup\u003e[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]\u003c/sup\u003e. These patterns served as controlled ground truth for comparing volume rendering and motion capture between 4D-CT and MRI.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section3\"\u003e \u003ch2\u003e2.2.2 Phantom Imaging Protocol\u003c/h2\u003e \u003cp\u003e \u003cstrong\u003e4D-CT\u003c/strong\u003e \u003cp\u003eThe CT imaging data from the motion phantom were acquired using the Siemens CT Big Bore (Siemens Healthineers, Erlangen, Germany) and saved in DICOM format. The scan parameters were determined based on visual assessment of the tumors in the CT images.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003e3D and 4D-MRI\u003c/b\u003e: A simulated 4D-MRI dataset was generated for the phantom by acquiring five static 3D T2W-MRI volumes at predefined positions across one motion cycle. Detailed acquisition parameters are provided in Additional file 1: Additional Methods.\u003c/p\u003e \u003cp\u003e \u003cb\u003eCine-MRI\u003c/b\u003e: Sagittal cine-MRI was acquired for the phantom using the default thorax protocol. Detailed sequence parameters are provided in Additional file 1: Additional Methods.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Patient Study\u003c/h2\u003e \u003cdiv id=\"Sec8\" class=\"Section3\"\u003e \u003ch2\u003e2.3.1 Patient Cohort\u003c/h2\u003e \u003cp\u003e Patients were prospectively enrolled (n\u0026thinsp;=\u0026thinsp;20; 10 lung and 10 liver) with histologically confirmed primary/metastatic tumors planned for SBRT, eligible for both 4D-CT and MR-Linac imaging, and age\u0026thinsp;\u0026ge;\u0026thinsp;18 years; all provided written informed consent. 4D-CT was routine simulation, and additional MR-Linac imaging (T2W-MRI and cine-MRI) was acquired under a research protocol prior to fraction 1 (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePatient and Tumor Characteristics\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\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 \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCharacteristic\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAll Patients (n\u0026thinsp;=\u0026thinsp;20)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eLung Cancer (n\u0026thinsp;=\u0026thinsp;10)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eLiver Cancer (n\u0026thinsp;=\u0026thinsp;10)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDemographics\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 \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\u003eMedian (Range) [59.9 (44.2\u0026ndash;75.3)]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMedian (Range) [62.4 (54.5\u0026ndash;75.3)]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMedian (Range) [57.4 (44.2\u0026ndash;66.7)]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGender\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMale (n) /Female (n) [13/7]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMale (n) / Female (n) [6/4]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMale (n) / Female (n) [7/3]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTumor Characteristics\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 \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePathology\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ePrimary / Metastatic [16/4]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eAdenocarcinoma (7), Squamous cell carcinoma (3)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eHCC (6), Metastasis (4)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eClinical Stage (AJCC 8th)\u003csup\u003e\u0026dagger;\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eI (4), II (10), III (4), IV (2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eI (8), II (2)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eI (4), II (1), III (3), IV (2)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGTV Volume on 4D-CT, cm\u0026sup3;\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMedian (Range) [31.2 (2.4-142.6)]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMedian (Range) [9.7 (2.4\u0026ndash;15.1)]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMedian (Range) [52.4 (5.5-142.6)]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMotion Amplitude (SI), mm*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD [18.5\u0026thinsp;\u0026plusmn;\u0026thinsp;4.7]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD [16.7\u0026thinsp;\u0026plusmn;\u0026thinsp;4.2]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD [20.2\u0026thinsp;\u0026plusmn;\u0026thinsp;5.6]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e \u003cp\u003eAbbreviations: HCC, hepatocellular carcinoma; SI, superior\u0026ndash;inferior; SD, standard deviation.\u003c/p\u003e \u003cp\u003e\u0026dagger; Clinical stage was assigned according to the corresponding disease-specific staging system used at the participating institution.\u003c/p\u003e \u003cp\u003e* Tumor motion amplitude was measured as peak-to-peak displacement in the superior\u0026ndash;inferior direction on 4D-CT.\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=\"Sec9\" class=\"Section3\"\u003e \u003ch2\u003e2.3.2 Patient Imaging Protocol\u003c/h2\u003e \u003cp\u003e \u003cstrong\u003e4D-CT Data Acquisition\u003c/strong\u003e \u003cp\u003ePatient 4D-CT and 3D-CT data were acquired on a Siemens SOMATOM Confidence CT Big Bore scanner (Siemens Healthineers, Erlangen, Germany) at 120 kVp and reconstructed into 10 respiratory phases with 1.5-mm slice thickness, 1.0-mm increment, and a B30f convolution kernel. Respiratory traces were recorded using the Real-time Position Management system (Varian Medical Systems, Palo Alto, CA, USA).\u003c/p\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e\n\u003cdiv class=\"Heading\"\u003e\u003cb\u003e3D MRI\u003c/b\u003e: Patients were imaged on an MR-Linac (T2W-MRI sequence).\u003c/div\u003e \u003cp\u003e \u003cstrong\u003e4D-MRI\u003c/strong\u003e \u003cp\u003e4D-MRI was not acquired in the patient cohort due to technical and workflow constraints at the time of the study, including the lack of a clinically streamlined 4D-MRI sequence on the MR-Linac and the extended scan time which was prohibitive within the clinical workflow.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003eCine MRI\u003c/b\u003e: During patient positioning, sagittal cine-MRI was acquired on the Unity MR-Linac (Elekta AB, Stockholm, Sweden) using the integrated 1.5T wide-bore MRI system and an 8-channel body array coil. Key parameters were TR/TE\u0026thinsp;=\u0026thinsp;3.6/1.5 ms and a temporal resolution of 0.5 s per frame; additional acquisition details are provided in Additional file 1: Additional Methods.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section3\"\u003e \u003cdiv class=\"Heading\"\u003e2.3.3 Contouring for Phantom and Patients\u003c/div\u003e \u003cp\u003eImage data were imported into MIM Maestro Workstation (version 7.1.2, MIM Software Inc., Cleveland, OH, USA) for GTV contouring. Six physicians independently delineated GTVs on each image set using a standardized contouring guideline, including fixed zoom and window/level settings (\u0026minus;\u0026thinsp;1024/300 HU for 4D-CT and 3020/1510 for T2W-MRI). For patient studies, only the primary tumor was contoured, and the ITV was defined as the union of all phase-specific 4D-CT GTVs without an additional margin. In addition, two senior radiation oncologists independently assessed tumor boundary clarity on co-registered 4D-CT and T2W-MRI and reached a consensus for each case.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Data and Statistical Analysis\u003c/h2\u003e \u003cdiv id=\"Sec13\" class=\"Section3\"\u003e \u003ch2\u003e2.4.1 Volumetric and Shape Comparison Metrics\u003c/h2\u003e \u003cp\u003eVolume metrics, the Dice Similarity Coefficient (DSC)\u003csup\u003e[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e],\u003c/sup\u003e and the 95th percentile Hausdorff Distance (HD95) were used to compare delineations.\u003c/p\u003e \u003cp\u003e \u003cstrong\u003eDice Similarity Coefficient (DSC)\u003c/strong\u003e \u003cp\u003eThe DSC quantifies the volumetric overlap between two contours, A and B. It is defined as DSC(A, B)\u0026thinsp;=\u0026thinsp;2|A \u0026cap; B| / (|A| + |B|), where |\u0026middot;| denotes volume. DSC values range from 0 (no overlap) to 1 (perfect overlap).\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eHausdorff Distance (HD95)\u003c/strong\u003e \u003cp\u003eThe Hausdorff Distance measures the maximum distance between two contours. To reduce sensitivity to single outlier points, we employed the 95th percentile Hausdorff Distance (HD95)\u003csup\u003e[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]\u003c/sup\u003e, which represents the 95th percentile of all distances between points on contour A and the nearest point on contour B (and vice versa), rather than the absolute maximum. A smaller HD95 indicates better geometric agreement.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eCoefficient of Variation (CV)\u003c/strong\u003e \u003cp\u003eThe CV was used to assess the relative volumetric variability of GTVs across respiratory phases. It was calculated as CV = (σ / \u0026micro;) \u0026times; 100%, where σ is the standard deviation and \u0026micro; is the mean GTV volume for a given tumor across phases.\u003c/p\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section3\"\u003e \u003ch2\u003e2.4.2 Motion Analysis\u003c/h2\u003e \u003cp\u003eTumor motion amplitude (peak-to-peak displacement in the primary direction) was measured from both 4D-CT and cine-MRI datasets. The correlation between motion amplitudes derived from these two modalities was assessed for both phantom and patient data.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section3\"\u003e \u003ch2\u003e2.4.3 Statistical analysis\u003c/h2\u003e \u003cp\u003eStatistical analyses were performed using SPSS 26.0 and R 4.2.2. Normality was assessed using the Shapiro\u0026ndash;Wilk test. Paired t-tests or Wilcoxon signed-rank tests were used for within-subject comparisons, and independent t-tests or Mann\u0026ndash;Whitney U tests were used for between-group comparisons, as appropriate. Categorical variables were compared using the Chi-square test or Fisher\u0026rsquo;s exact test. A two-sided p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was considered statistically significant.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Phantom Results\u003c/h2\u003e \u003cdiv id=\"Sec18\" class=\"Section3\"\u003e \u003ch2\u003e3.1.1 Phantom Volumetric Analysis\u003c/h2\u003e \u003cp\u003eThe phantom GTV volume results from 4D-CT are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. The average absolute volume of GTVs across the 10 phases was 31.16 cm\u0026sup3;, with a narrow 95% confidence interval (CI) ranging from 31.00 to 31.32 cm\u0026sup3;. The average standard deviation (SD) of GTVs across the five simulated motion modes was 0.3176 cm\u0026sup3;, with individual mode SDs ranging from 0.053 to 0.58 cm\u0026sup3;. The mean coefficient of variation for the phantom GTVs was 1.79%, with a range of 0.85% to 5.76%, demonstrating high volumetric consistency in the controlled phantom setup.\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\u003ePhantom GTV Volumes Across Motion Patterns and Imaging Modalities\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\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 \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"7\" nameend=\"c7\" namest=\"c1\"\u003e \u003cp\u003eA. Absolute GTV Volumes (cm\u0026sup3;) by Motion Pattern\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eImaging Modality\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSin-10A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSin-25A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCos4AP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eHeart\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eSawtooth\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eMean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4D-CT\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e52.41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e66.58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e54.39\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e45.28\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e51.42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e54.02\u0026thinsp;\u0026plusmn;\u0026thinsp;7.86\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3D-CT\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e51.88\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e65.61\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e52.87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e47.15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e50.53\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e53.61\u0026thinsp;\u0026plusmn;\u0026thinsp;7.55\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eT2W-MRI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e56.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e63.01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e66.77\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e48.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e56.46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e58.23\u0026thinsp;\u0026plusmn;\u0026thinsp;7.51\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4D-MRI\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e57.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e72.35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e55.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e46.54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e55.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e57.39\u0026thinsp;\u0026plusmn;\u0026thinsp;9.36\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"6\" nameend=\"c6\" namest=\"c1\"\u003e \u003cp\u003eB. Volumetric Variability Metrics for 4D-CT Across 10 Respiratory Phases\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMotion Pattern\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMean GTV (cm\u0026sup3;)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSD (cm\u0026sup3;)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCV (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e95% CI Lower\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e95% CI Upper\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSin-10A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e31.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.39\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e31.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e31.19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSin-25A\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e31.25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.58\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e31.15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e31.35\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCos4AP\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e31.08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e31.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e31.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHeart\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e31.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.85\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e30.9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e31.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSawtooth\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e31.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e31.07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e31.13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"7\" nameend=\"c7\" namest=\"c1\"\u003e \u003cp\u003eAbbreviations: GTV, gross tumor volume; SD, standard deviation; CV, coefficient of variation; CI, confidence interval; CT, computed tomography; MRI, magnetic resonance imaging.\u003c/p\u003e \u003cp\u003ePart A summarizes absolute GTV measurements across motion patterns and imaging modalities. Part B summarizes volumetric variability metrics derived from 4D-CT across 10 respiratory phases.\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eIn the motion phantom, the average GTV size from T2W-MRI across the five motion modes was 56.69 cm\u0026sup3; (range: 49.41\u0026ndash;67.04 cm\u0026sup3;). Similarly, the average GTV size from 4D-MRI was 57.39 cm\u0026sup3;, with a slightly wider range of 45.79 to 68.99 cm\u0026sup3;.\u003c/p\u003e \u003cp\u003eThe motion phantom was assessed using 4D-CT, 4D-MRI, and T2W-MRI. The results showed no significant difference (p\u0026thinsp;=\u0026thinsp;0.2506) between the average GTV obtained from 4D-MRI imaging and the results from 4D-CT. For the phantom, the median GTV sizes across the four imaging modalities were as follows: 3D-CT GTV (median: 51.88 cm\u0026sup3;, range: 47.15\u0026ndash;65.61 cm\u0026sup3;), T2W-MRI GTV (median: 56.69 cm\u0026sup3;, range: 48.20-66.77 cm\u0026sup3;), 4D-CT GTV (median: 52.40 cm\u0026sup3;, range: 44.32\u0026ndash;63.71 cm\u0026sup3;), and 4D-MRI GTV (median: 55.50 cm\u0026sup3;, range: 46.54\u0026ndash;72.35 cm\u0026sup3;).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section3\"\u003e \u003ch2\u003e3.1.2 Phantom Motion Analysis\u003c/h2\u003e \u003cp\u003eBased on the phantom results, 4D-CT may slightly underestimate the full motion envelope relative to MR-based techniques, as summarized in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eB and Additional file 1: Table S2.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Patient Results\u003c/h2\u003e \u003cdiv id=\"Sec21\" class=\"Section3\"\u003e \u003ch2\u003e3.2.1 Patient Volumetric and Shape Analysis\u003c/h2\u003e \u003cp\u003eGTV Volume and Variability on 4D-CT: The detailed results for GTV volume are presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003e. As illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003eA, the average standard deviation (SD) of GTVs from the 10 phases was 1.54 cm\u0026sup3; (range: 1.02\u0026ndash;3.45 cm\u0026sup3;) for lung patients. For liver patients, the average SD was notably higher at 3.66 cm\u0026sup3;, ranging from 2.58 to 5.68 cm\u0026sup3;. Figure\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e2\u003c/span\u003eB displays the coefficients of variation (CV) for GTV volumes across respiratory phases. Liver tumors demonstrated significantly greater volumetric variability during respiration than lung tumors (median CV: 8.95% vs 5.72%; Mann\u0026ndash;Whitney U test, p\u0026thinsp;=\u0026thinsp;0.0211).\u003c/p\u003e \u003cp\u003eThe volumetric variations of GTVs at different respiratory phases (GTV-0%, GTV-20%, GTV-50%, and GTV-70%) are detailed in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003eA. For liver patients, the mean GTV sizes were 50.63 cm\u0026sup3;, 52.35 cm\u0026sup3;, 55.78 cm\u0026sup3;, and 53.76 cm\u0026sup3;, respectively. For lung patients, the corresponding mean sizes were 7.54 cm\u0026sup3;, 8.25 cm\u0026sup3;, 9.27 cm\u0026sup3;, and 8.15 cm\u0026sup3;. Statistical analysis revealed that the volumetric variation among the 10 phases was insignificant (F\u0026thinsp;=\u0026thinsp;0.005, P\u0026thinsp;=\u0026thinsp;0.999).\u003c/p\u003e \u003cp\u003eHowever, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e3\u003c/span\u003eB, the 3D-CT GTV size in lung patients (7.89\u0026thinsp;\u0026plusmn;\u0026thinsp;4.99 cm\u0026sup3;) was significantly smaller than the 4D-CT size (9.71\u0026thinsp;\u0026plusmn;\u0026thinsp;6.04 cm\u0026sup3;) (P\u0026thinsp;=\u0026thinsp;0.001). A similar significant difference was observed in liver patients, where the 3D-CT GTV size (50.63\u0026thinsp;\u0026plusmn;\u0026thinsp;61.90 cm\u0026sup3;) was smaller than the 4D-CT size (69.83\u0026thinsp;\u0026plusmn;\u0026thinsp;66.57 cm\u0026sup3;) (P\u0026thinsp;=\u0026thinsp;0.001).\u003c/p\u003e \u003cp\u003eComparison between MRI and CT-based GTVs: A comparative analysis of volumetric variation was performed on ten cases of lung and liver tumors preoperatively evaluated with 4D-CT, 3D-CT, and T2W-MRI imaging, as visualized in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eA. A statistically significant difference (p\u0026thinsp;=\u0026thinsp;0.0285) was observed between the average GTV obtained from MR imaging and the results from 4D-CT for both lung and liver patients. The inter-observer variability, reflected in the range of individual GTV volumes contoured by the six physicians, showed a pooled standard deviation of 1.8 cm\u0026sup3; for CT-based contours and 2.3 cm\u0026sup3; for MRI-based contours across the patient cohort. Qualitative assessment indicated that MR imaging demonstrated clearer and more precise tumor contours compared to contrast-enhanced CT. Furthermore, the data range illustrated in the figures clearly indicates that the volumetric range captured by 4D-CT is more extensive than that captured by T2W-MRI.\u003c/p\u003e \u003cp\u003eThe size of the GTV is significantly influenced by motion and delineation uncertainties, particularly for small mobile tumors. The relative difference in GTV, calculated as [(GTV₃DMR - GTV₃DCT) / GTV₃DCT \u0026times; 100%], ranged from 2.2% to 26.2%, in absolute terms, this corresponded to volume differences ranging from 0.8 cm\u0026sup3; to 18.5 cm\u0026sup3;, with the underlying 3D-CT GTVs ranging from 2.4 to 142.6 cm\u0026sup3;, as detailed in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eA. This represents a clinically significant variation in target volume for radiotherapy treatment planning. Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e further presents the HD95 and DICE values quantifying the inter-modality shape agreement for GTVs delineated by six physicians between 3D-CT and T2W-MRI. The average HD95 was 2.9\u0026thinsp;\u0026plusmn;\u0026thinsp;1.5 mm, and the mean Dice Similarity Coefficient (DSC) was 0.72\u0026thinsp;\u0026plusmn;\u0026thinsp;0.11.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eInter-modality Agreement Between 3D-CT and T2W-MRI\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"8\"\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 \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSin-10A\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSin-25A\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eCos4AP\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eHeart\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eSawtooth\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eLiver\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eLung\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDSC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.73\u0026thinsp;\u0026plusmn;\u0026thinsp;0.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.63\u0026thinsp;\u0026plusmn;\u0026thinsp;0.18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.77\u0026thinsp;\u0026plusmn;\u0026thinsp;0.16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.61\u0026thinsp;\u0026plusmn;\u0026thinsp;0.17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.75\u0026thinsp;\u0026plusmn;\u0026thinsp;0.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.59\u0026thinsp;\u0026plusmn;\u0026thinsp;0.21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.73\u0026thinsp;\u0026plusmn;\u0026thinsp;0.15\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e(mean\u0026thinsp;\u0026plusmn;\u0026thinsp;std)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHD95(mm)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e2.26\u0026thinsp;\u0026plusmn;\u0026thinsp;1.2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e4.26\u0026thinsp;\u0026plusmn;\u0026thinsp;1.8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e2.12\u0026thinsp;\u0026plusmn;\u0026thinsp;1.1\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e5.54\u0026thinsp;\u0026plusmn;\u0026thinsp;1.5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e1.98\u0026thinsp;\u0026plusmn;\u0026thinsp;1.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e6.35\u0026thinsp;\u0026plusmn;\u0026thinsp;2.3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e2.34\u0026thinsp;\u0026plusmn;\u0026thinsp;1.2\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e(mean\u0026thinsp;\u0026plusmn;\u0026thinsp;std)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colspan=\"8\" nameend=\"c8\" namest=\"c1\"\u003e \u003cp\u003eAbbreviations: DSC, Dice similarity coefficient; HD95, 95th percentile Hausdorff distance; CT, computed tomography; MRI, magnetic resonance imaging; SD, standard deviation.\u003c/p\u003e \u003cp\u003eValues are presented as mean\u0026thinsp;\u0026plusmn;\u0026thinsp;SD. Phantom columns summarize inter-modality agreement across predefined motion patterns, whereas Liver and Lung summarize patient-based comparisons between 3D-CT and T2-weighted MRI.\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=\"Sec22\" class=\"Section3\"\u003e \u003ch2\u003e3.2.2 Patient Motion Analysis\u003c/h2\u003e \u003cp\u003eTumor motion amplitudes measured by cine-MRI were strongly correlated with those measured by 4D-CT, as presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e5\u003c/span\u003eB and Additional file 1: Table S2. Nevertheless, real-time cine-MRI occasionally revealed motion excursions beyond the 4D-CT-based internal target volume.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec23\" class=\"Section2\"\u003e \u003ch2\u003e3.3 Representative Image Comparison\u003c/h2\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e4\u003c/span\u003e depicts a representative example of the difference in image quality between 4D-CT and 4D-MRI of the phantoms. This qualitative disparity can directly impact the accuracy of tumor delineation, especially for smaller tumors with substantial motion. The observed variation in GTV within the breathing cycle may be attributed to artifacts inherent in 4D image quality and intra-observer variation during delineation.\u003c/p\u003e \u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003eIn radiation physics, dynamic phantom studies are widely employed to verify target volume and motion variations across different imaging modalities, utilizing predefined volumes, motion patterns, and amplitudes for assessment and comparison\u003csup\u003e[\u003cspan additionalcitationids=\"CR20 CR21\" citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]\u003c/sup\u003e. While phantoms simulate motion under idealized conditions, it is crucial to recognize that their results may diverge from those observed in actual patient studies. Internationally, several studies have focused on the development of dynamic MRI phantoms\u003csup\u003e[\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e, \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn this study, we conducted a comprehensive investigation using a dynamic thorax phantom compatible with both MRI and CT, complemented by clinical studies on patients with lung and liver tumors. Our primary objective was to collect clinical data on tumor motion and volume variation using 4D-CT, 4D-MRI, and real-time cine MRI. It is important to clarify that 4D MRI was not acquired in the patient cohort due to technical and workflow constraints at the time of the study (e.g., lack of a clinically streamlined sequence and prohibitive scan length). Therefore, our conclusions regarding 4D-MRI are derived from the phantom experiment, while the patient study focuses on comparing the primary MRI sequence used for online delineation (T2W-MRI) with the standard simulation modality (4D-CT). An observed phenomenon in our study was the systematically larger GTVs delineated on T2W-MRI compared to 4D-CT. This discrepancy can be partly understood by considering how motion impacts these modalities differently. A free-breathing T2W-MRI, with its longer acquisition time, produces a time-averaged intensity image of the tumor across many breathing cycles. The contoured GTV thus represents an intensity-averaged volume that inherently incorporates information about the tumor's shape over its motion path. In contrast, a GTV on a single-phase 4D-CT is a spatial snapshot at one moment, and even the ITV is a spatial union of such snapshots, not an intensity average. This fundamental difference in acquisition physics contributes to the systematic volume differences observed. A potential contributing factor to this discrepancy may be peritumoral edema, a well-documented imaging characteristic that is often more conspicuous on T2-weighted sequences due to their inherent sensitivity to fluid content, compared to CT\u003csup\u003e[\u003cspan additionalcitationids=\"CR26\" citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]\u003c/sup\u003e. This established principle offers a plausible explanation for this particular volumetric discrepancy, though it requires further direct histopathological correlation for confirmation in our cohort.\u003c/p\u003e \u003cp\u003eIn the phantom study, the target was delineated using both CT and T2-weighted MRI, with both modalities clearly displaying the target boundary, thereby confirming the reliability of the phantom model. However, in patient studies, a significant source of uncertainty in Gross Tumor Volume (GTV) delineation arises from the variable visualization of tumor boundaries, particularly between central and peripheral lesions\u003csup\u003e[\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]\u003c/sup\u003e Peripheral lesions often have well-defined boundaries, whereas central lesions may be attached to normal hilar structures, making their margins indistinct. While CT often struggles to delineate the boundaries of central lung cancers, the superior soft tissue contrast of T2-weighted MRI provides clearer definition, underscoring its clinical value.\u003c/p\u003e \u003cp\u003eWe further evaluated inter-observer variation using the Dice similarity coefficient (DSC) and 95th percentile Hausdorff distance (HD95) among six physicians contouring GTVs on MRI and CT across different respiratory patterns. A clear discrepancy in target volume between MRI and CT was observed in the phantom, particularly under motion patterns with larger amplitudes. A similar trend was also seen in patients, especially in liver tumors with greater respiratory motion. These findings indicate that apparent tumor boundaries are highly motion-dependent, with more pronounced MRI\u0026ndash;CT differences in liver tumors. This is consistent with Case et al., who reported smaller motion amplitudes on 4D-MRI than on 4D-CT \u003csup\u003e[\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]\u003c/sup\u003e. Therefore, multimodal imaging integration is essential for more accurate target delineation.\u003c/p\u003e \u003cp\u003eBreathing irregularities can introduce variations in tumor displacement throughout the breathing cycle during both 4D-CT and 4D-MRI acquisitions. For 4D-CT, the tumor motion is captured over a few bed positions or helical pitches, effectively representing an average over several breathing cycles\u003csup\u003e[\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]\u003c/sup\u003e. For 4D-MRI, the initial phase (typically around 10 seconds, covering 2\u0026ndash;3 cycles) determines the breathing amplitude used for the entire scan\u003csup\u003e[\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]\u003c/sup\u003e. Consequently, the Internal Target Volume (ITV) derived from either 4D technique may not fully capture the tumor motion over extended periods, such as during a radiotherapy fraction. Investigating the discrepancies in tumor motion between 4D-CT and 4D-MRI, and evaluating the differences between simulated localization and real-time measurement, remain topics of active research\u003csup\u003e[\u003cspan additionalcitationids=\"CR34\" citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThis study used cine MRI on the MR-Linac to address this issue. Phantom and patient data showed that tumor motion amplitude measured on simulation 4D-CT better matched real-time motion during radiotherapy. Phantom results also confirmed that 4D-MRI-based simulated localization accurately represented tumor motion amplitude, supporting its clinical value.\u003c/p\u003e \u003cp\u003eIn some patients, cine MRI detected motion beyond the ITV from simulation 4D-CT, likely because the brief 4D scan failed to capture irregular breathing and complex motion. In such cases, adaptive MR-Linac replanning may be needed to maintain accuracy. For motion-managed radiotherapy, SBRT with real-time target monitoring is strongly recommended.\u003c/p\u003e \u003cp\u003eA key finding of this study is the significant disparity in GTV volume variability between phantom and patient 4D-CT images. The coefficients of variation for lung and liver tumor GTVs in patients were 5.72% and 8.95%, respectively, substantially larger than the 1.79% variation observed in the phantom. These results indicate that real tumor motion and morphological changes during respiration are far more complex and must be carefully accounted for in clinical studies and practice. The decision to utilize a 1.5T MR-Linac in this study was informed by prior evidence suggesting that, for thoracic applications, 1.5T systems may offer advantages in mitigating respiratory motion artifacts and geometric distortion compared to 3.0T systems \u003csup\u003e[\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThis study has several limitations. The primary limitation was the relatively small patient sample size, which may constrain the generalizability of the findings and the statistical power for subgroup analyses. All within-subject comparisons between imaging modalities were re-analyzed using appropriate paired statistical tests to ensure robust inference. A key technical limitation was the inability to acquire patient 4D-MRI due to workflow and sequence availability constraints at the time of the study, restricting our comparative 4D-MRI analysis to the controlled phantom environment. Finally, the patient cohort was selected based on suitability for both CT and MRI, which may introduce a selection bias by excluding tumors poorly visualized on either modality.\u003c/p\u003e"},{"header":"5. Conclusions","content":"\u003cp\u003e T2W-MRI provides superior soft-tissue contrast for GTV delineation and may improve MRI-based target definition in MR-guided radiotherapy. While 4D-CT and phantom-based 4D-MRI provide reasonable motion estimates during simulation, they may not fully capture the entirety of real-time tumor motion during treatment. Real-time cine-MRI on the MR-Linac therefore provides important complementary information for adaptive SBRT. Integrating CT-based motion assessment with MRI-based soft-tissue visualization may further improve the precision of SBRT delivery.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003e4D-CT \u0026nbsp; \u0026nbsp; \u0026nbsp;four-dimensional computed tomography\u003c/p\u003e\n\u003cp\u003e4D-MRI \u0026nbsp; \u0026nbsp;4D magnetic resonance imaging\u003c/p\u003e\n\u003cp\u003eAP \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; anterior-posterior\u003c/p\u003e\n\u003cp\u003eFSE \u0026nbsp; \u0026nbsp; \u0026nbsp;Fast spin echo\u003c/p\u003e\n\u003cp\u003eDSC \u0026nbsp; \u0026nbsp; Dice Similarity Coefficient\u003c/p\u003e\n\u003cp\u003eHD95 \u0026nbsp; Hausdorff distance\u003c/p\u003e\n\u003cp\u003eITV \u0026nbsp; \u0026nbsp; \u0026nbsp; internal target volume\u003c/p\u003e\n\u003cp\u003eMRI \u0026nbsp; \u0026nbsp; \u0026nbsp;magnetic resonance imaging\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eNSA \u0026nbsp; \u0026nbsp; Number of Signal Average\u003c/p\u003e\n\u003cp\u003eOAR \u0026nbsp; \u0026nbsp; organs at risk\u003c/p\u003e\n\u003cp\u003eSI \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; superior-inferior\u003c/p\u003e\n\u003cp\u003eSBRT \u0026nbsp; \u0026nbsp;Stereotactic body radiation therapy\u003c/p\u003e\n\u003cp\u003eTR \u0026nbsp; \u0026nbsp; \u0026nbsp;repetition time \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTE \u0026nbsp; \u0026nbsp; \u0026nbsp;echo time \u0026nbsp;\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was reviewed and approved by the Ethics Committee of Shandong First Medical University Affiliated Cancer Hospital (Approval No. SDTHEC-202409048) and was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets generated and/or analyzed during the current study are not publicly available due to patient privacy and institutional restrictions but are available from the corresponding author on reasonable request and with permission from the participating institution(s).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported in part by the National Natural Science Foundation of China (Grant No. 12475343), the Collaborative Academic Innovation Project of Shandong Cancer Hospital (TS004), and the Taishan Scholars Program of Shandong Province, China (Grant No. ts202408375).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors' contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eJHC, HCY, and LM conceived and designed the study. JHC, HCY, LM, XA, XT, GL, and TH collected the data and performed image analysis and contouring. HCY, YY, XCL, and ZJL contributed to methodology development, data interpretation, and supervision. JHC drafted the manuscript. XCL and ZJL critically revised the manuscript for important intellectual content. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors gratefully acknowledge all patients who participated in this study and the clinical staff who supported image acquisition and data collection.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors' information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eFlentje M. Stereotactic Radiotherapy of Targets in the Lung and Liver. Strahlenther Onkol. 2001;177:645\u0026ndash;55.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNavarro-Martin A, Aso S, Cacicedo J, et al. Phase II Trial of SBRT for Stage I NSCLC: Survival, Local Control, and Lung Function at 36 Months. J Thorac Oncol. 2016;11:1101\u0026ndash;11.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBorm KJ, Oechsner M, Wiegandt M, Hofmeister A, Combs SE, Duma MN. Moving targets in 4D-CTs versus MIP and AIP: comparison of patients data to phantom data. BMC Cancer. 2018;18.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNakamura M, Narita Y, Matsuo Y, et al. Geometrical differences in target volumes between slow CT and 4D CT imaging in stereotactic body radiotherapy for lung tumors in the upper and middle lobe. Med Phys. 2008;35:4142\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRietzel E, Liu AK, Doppke KP, et al. Design of 4D treatment planning target volumes. Int J Radiat Oncol Biol Phys. 2006;66:287\u0026ndash;95.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBradley JD, Nofal AN, Naqa IME, et al. Comparison of helical, maximum intensity projection (MIP), and averaged intensity (AI) 4D CT imaging for stereotactic body radiation therapy (SBRT) planning in lung cancer. Radiotherapy Oncol. 2006;81:264\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRebecca, Muirhead S, et al. Use of Maximum Intensity Projections (MIPs) for Target Outlining in 4D-CT Radiotherapy Planning. J Thorac Oncol. 2008;3:1433\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXue H, Artico J, Fontana M, Moon JC, Davies RH, Kellman P. Landmark Detection in Cardiac MRI by Using a Convolutional Neural Network. Radiol Artif Intell. 2021;3(5):e200197.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKeall PJ, Mageras GS, Balter JM, Emery RS, Forster KM, Jiang SB, Kapatoes JM, Low DA, Murphy MJ, Murray BR, Ramsey CR, Van Herk MB, Vedam SS, Wong JW, Yorke E. The management of respiratory motion in radiation oncology report of AAPM Task Group 76. Med Phys. 2006;33(10):3874\u0026thinsp;\u0026ndash;\u0026thinsp;900. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1118/1.2349696\u003c/span\u003e\u003cspan address=\"10.1118/1.2349696\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 17089851.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCase RB, Sonke JJ, Moseley DJ, Kim J, Brock KK, Dawson LA. Inter- and intrafraction variability in liver position in non-breath-hold stereotactic body radiotherapy. Int J Radiat Oncol Biol Phys. 2009;75(1):302\u0026ndash;8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ijrobp.2009.03.058\u003c/span\u003e\u003cspan address=\"10.1016/j.ijrobp.2009.03.058\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2009 Jul 21. PMID: 19628342.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBourque AE, Bedwani S, Filion D, Carrier JF. A particle filter based autocontouring algorithm for lung tumor tracking using dynamic magnetic resonance imaging. Med Phys. 2016;43:5161\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIpsen S, Blanck O, Lowther NJ, et al. Towards real-time MRI-guided 3D localization of deforming targets for non-invasive cardiac radiosurgery. Phys Med Biol. 2016;61:7848.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBourque AE, Bedwani S, Carrier JF, et al. Particle Filter-Based Target Tracking Algorithm for Magnetic Resonance-Guided Respiratory Compensation: Robustness and Accuracy Assessment. Int J Radiat Oncol Biol Phys. 2018;100:325\u0026ndash;34.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKeall PJ, Mageras GS, Balter JM, Emery RS, Yorke E. The management of respiratory motion in radiation oncology report of AAPM Task Group 76. Med Phys. Med Phys. 2006;33:3874\u0026ndash;900.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShimizu S, Shirato H, Ogura S, et al. Detection of lung tumor movement in real-time tumor-tracking radiotherapy. Int J Radiat Oncol Biol Phys. 2001;51:304\u0026ndash;10.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSpautz S, Haase L, Tschiche M, Makocki S, Richter C, Troost EGC, St\u0026uuml;tzer K. Comparison of 3D and 4D robustly optimized proton treatment plans for non-small cell lung cancer patients with tumour motion amplitudes larger than 5 mm. Phys Imaging Radiat Oncol. 2023;27:100465. PMID: 37449022; PMCID: PMC10338142.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZou KH, Warfield SK, Bharatha A, Tempany CM, Kaus MR, Haker SJ, Wells WM 3rd, Jolesz FA, Kikinis R. Statistical validation of image segmentation quality based on a spatial overlap index. Acad Radiol. 2004;11(2):178\u0026ndash;89. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/s1076-6332(03)00671-8\u003c/span\u003e\u003cspan address=\"10.1016/s1076-6332(03)00671-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. PMID: 14974593; PMCID: PMC1415224.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuttenlocher DP, Gregory A, Klanderman. Rucklidge. Comparing images using the Hausdorff distance. IEEE Trans Pattern Anal Mach Intell. 2002;15(9):850\u0026ndash;63.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWeidner A, Stengl C, Dinkel F, et al. An abdominal phantom with anthropomorphic organ motion and multimodal imaging contrast for MR-guided radiotherapy. Phys Med Biol. 2022;67:67.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNiebuhr NI, Johnen W, Echner G, et al. The ADAM-pelvis phantom-an anthropomorphic, deformable and multimodal phantom for MRgRT. Phys Med Biol. 2019;64:04NT05.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNikolai J, Mickevicius X et al. Simultaneous motion monitoring and truth-in-delivery analysis imaging framework for MR-guided radiotherapy. Phys Med Biol. 2018.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBowen SR, Nyflot MJ, Herrmann C, et al. Imaging and dosimetric errors in 4D PET/CT-guided radiotherapy from patient-specific respiratory patterns: a dynamic motion phantom end-to-end study. Phys Med Biol. 2015;60:3731\u0026ndash;46.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKeizer DMDM, Kerkmeijer LGW, Willigenburg T, Lier ALHMWv, Boer JCJD. Prostate intrafraction motion during the preparation and delivery of MR-guided radiotherapy sessions on a 1.5T MR-Linac. Radiother Oncol 2020;151.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSergej, Schneider K et al. Commissioning of a 4D MRI phantom for use in MR-guided radiotherapy. Med Phys. 2018.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoy AE, Wells P. Volume definition in radiotherapy planning for lung cancer: how the radiologist can help. Cancer imaging: official publication Int Cancer Imaging Soc. 2006;6:116\u0026ndash;23.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLampen-Sachar K, Zhao B, Zheng J, et al. Correlation between tumor measurement on Computed Tomography and resected specimen size in lung adenocarcinomas. Lung Cancer. 2012;75:332\u0026ndash;5.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYu HM, Liu YF, Hou M, Liu J, Li XN, Yu JM. Evaluation of gross tumor size using CT, 18F-FDG PET, integrated 18F-FDG PET/CT and pathological analysis in non-small cell lung cancer. Eur J Radiol. 2009;72:104\u0026ndash;13.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNestle U, De Ruysscher D, Ricardi U, et al. ESTRO ACROP guidelines for target volume definition in the treatment of locally advanced non-small cell lung cancer. Radiotherapy oncology: J Eur Soc Therapeutic Radiol Oncol. 2018;127:1\u0026ndash;5.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCase RB, Sonke JJ, Moseley DJ, Kim J, Brock KK, Dawson LA. Inter- and intrafraction variability in liver position in non-breath-hold stereotactic body radiotherapy. Int J Radiat Oncol Biol Phys. 2009;75:302\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evan de Lindt TN, Nowee ME, Janssen T, et al. Technical feasibility and clinical evaluation of 4D-MRI guided liver SBRT on the MR-Linac. Radiother Oncol. 2022;167:285\u0026ndash;91.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStevens CW, Munden RF, Forster KM, et al. Respiratory-driven lung tumor motion is independent of tumor size, tumor location, and pulmonary function. Int J Radiat Oncol Biol Phys. 2001;51:62\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang J, Srivastava S, Wang C, et al. Clinical evaluation of 4D MRI in the delineation of gross and internal tumor volumes in comparison with 4D-CT. J Appl Clin Med Phys. 2019;20:51\u0026ndash;60.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi G, Citrin D, Camphausen K, Mueller B, Burman C, Mychalczak B, Miller RW, Song Y. Advances in 4D medical imaging and 4D radiation therapy. Technol Cancer Res Treat. 2008;7(1):67\u0026ndash;81.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evon Siebenthal M, Szekely G, Gamper U, Boesiger P, Lomax A, Cattin P. 4D MR imaging of respiratory organ motion and its variability. Phys Med Biol. 2007;52:1547\u0026ndash;64.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSlagowski JM, Ding Y, Aima M, et al. A modular phantom and software to characterize 3D geometric distortion in MRI. Phys Med Biol. 2020;65:195008.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu X, Li Z, Rong Y, Cao M, Qiu J. A Comparison of the Distortion in the Same Field MRI and MR-Linac System With a 3D Printed Phantom. Front Oncol 2021;11.\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":"radiation-oncology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"raon","sideBox":"Learn more about [Radiation Oncology](http://ro-journal.biomedcentral.com/)","snPcode":"13014","submissionUrl":"https://submission.nature.com/new-submission/13014/3","title":"Radiation Oncology","twitterHandle":"@OncoBioMed","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Volume variation, 4D-MRI, cine-MRI, Moving targets, MR-Linac","lastPublishedDoi":"10.21203/rs.3.rs-9157737/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9157737/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003ePrecise delineation of gross tumor volume (GTV) and accurate assessment of respiratory motion are critical for stereotactic body radiotherapy (SBRT) of lung and liver tumors. This study aimed to systematically compare CT- and MRI-based tumor volume delineation and motion assessment using a dynamic phantom and patient data, with phantom-based 4D-MRI and real-time cine-MRI included alongside standard 4D-CT.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eA 4D dynamic thorax phantom and twenty patients (ten lung, ten liver) were imaged using 4D-CT and MR-Linac MRI. Patients underwent T2W-MRI and cine-MRI, while a simulated 4D-MRI dataset was generated for the phantom only. Gross tumor volumes (GTVs) were contoured by six physicians on different image sets. Volume metrics, Dice Similarity Coefficient (DSC), and Hausdorff Distance (HD95) were used to compare delineations. Tumor motion amplitude from 4D-CT and cine-MRI was correlated.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eFor patients, the GTV derived from T2W-MRI was significantly larger than that from 4D-CT (p\u0026thinsp;=\u0026thinsp;0.028), with inter-modality delineation variability of DSC\u0026thinsp;=\u0026thinsp;0.72\u0026thinsp;\u0026plusmn;\u0026thinsp;0.11 and HD95\u0026thinsp;=\u0026thinsp;2.9\u0026thinsp;\u0026plusmn;\u0026thinsp;1.5 mm. In the phantom, 4D-MRI and 4D-CT GTVs showed no significant difference (p\u0026thinsp;=\u0026thinsp;0.251). Tumor motion amplitudes measured by cine-MRI were strongly correlated with those measured by 4D-CT (r\u0026thinsp;\u0026gt;\u0026thinsp;0.9), and no significant overall difference was observed between the paired measurements (p\u0026thinsp;=\u0026thinsp;0.626). However, real-time cine-MRI occasionally revealed motion excursions beyond the 4D-CT-based internal target volume.\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eT2W-MRI provides superior soft-tissue contrast for GTV delineation, often revealing a larger, potentially more accurate tumor volume. While 4D-CT and phantom-based 4D-MRI provide comparable motion estimates, real-time cine-MRI on the MR-Linac is essential to capture the full extent of tumor motion and guide adaptive SBRT for optimal accuracy.\u003c/p\u003e","manuscriptTitle":"Discrepancies in Gross Tumor Volume and Motion between 4D-CT and Magnetic Resonance Imaging for Stereotactic Body Radiotherapy: A Phantom and Patient Study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-20 10:20:15","doi":"10.21203/rs.3.rs-9157737/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"324878458460018984845531196218592138292","date":"2026-04-28T16:13:11+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-04-10T11:00:00+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-30T05:03:48+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-27T08:44:13+00:00","index":"","fulltext":""},{"type":"submitted","content":"Radiation Oncology","date":"2026-03-26T08:45:47+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"radiation-oncology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"raon","sideBox":"Learn more about [Radiation Oncology](http://ro-journal.biomedcentral.com/)","snPcode":"13014","submissionUrl":"https://submission.nature.com/new-submission/13014/3","title":"Radiation Oncology","twitterHandle":"@OncoBioMed","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"fca6c9a3-815c-452d-a873-0effa65a9a02","owner":[],"postedDate":"April 20th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-04-20T10:20:15+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-20 10:20:15","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9157737","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9157737","identity":"rs-9157737","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}