RPS 1815 - Predictive power of vascular imaging

Author Block: X. Han¹, X. Yu², Y. Zhou¹; ¹shanghai/CN, ²Shanghai/CN
Purpose: To develop a prediction model for WMH progression and to investigate whether glucometabolic dysregulation causally influences WMH progression via microstructural damage.
Methods or Background: The study analyzed data from UK Biobank participants of European descent with serial brain MRI scans. Participants took part in both imaging visit and with a diagnosis of CSVD were included in the study. For machine learning, a total of 8 key features were identified from Akaike information criterion, including age, body mass index, cystatin C, glucose, fractional anisotropy, mean diffusivity, intracellular volume fraction, and isotropic volume fraction. For Mendelian Randomization, 4 glucose indexes including fasting plasma glucose, 2-hour plasma glucose after an oral glucose tolerance test, HbA1c, fasting insulin, and 75 diffusion MRI (dMRI)-derived microstructural metrics were applied. Machine learning models were constructed to predict WMH progression, while structural equation modeling tested mediation pathways. Bidirectional two-sample Mendelian Randomization was employed to establish causal relationships between glucose metabolism indices and white matter microstructure using genome-wide association study data.
Results or Findings: Of 1616 participants included, 902 had WMH increase and 714 had WMH. Seven algorithms were employed to develop WMH prediction models, with logistic regression and support vector machine (SVM) demonstrating optimal performance. Structural equation modeling (SEM) revealed that glucose partially mediates WMH progression through ISOVF. Mendelian randomization (MR) analyses indicated that genetic susceptibility to hemoglobin A1c (HbA1c) significantly altered free water content in the left cerebral peduncle, right hippocampal gyrus, left anterior thalamic radiation, and left corticospinal tract.
Conclusion: Glucometabolic dysregulation contributed to WMH progression via microstructural damage.
Limitations: our study population was predominantly of European cohort from UK Biobank, which may limit the generalizability of the findings to other ethnic groups.
Funding for this study: This work was supported by National Natural Science Foundation of China (82171885), Shanghai Natural Science Foundation (25ZR1401225), Eastern Talent Plan Leading Project (LJ2023127), the Shanghai Science and Technology Committee Project, Explorer Project Funding (24TS1414800), the Leading Talent Program of Shanghai Municipal Health Commission (2022LJ023), Shanghai
Engineering Research Center of Peri-operative Organ Support and Function Preservation (20DZ2254200), Renji Hospital Project (RJTJ25-QN-064, RJTJ23-RC-013, RJTJ25-MS-014, RJKY24-004).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Participant data were obtained from the UK Biobank cohort (Approved Application No. 117280). Ethical approval for the UK Biobank Study was granted by the National Information Governance Board for Health and Social Care and the NHS North West Multicentre Research Ethics Committee.

Author Block: G. Lin, W. Chen, M. Chen, J. Ji; Lishui/CN
Purpose: To evaluate the predictive value of CTA-based radiomics (Rad) and deep learning (DL), combined with clinical and imaging features, for two-year ischemic stroke risk in patients with asymptomatic carotid plaque.
Methods or Background: We retrospectively studied 528 asymptomatic patients who underwent CTA. Patients were randomly allocated to training (n = 370) and validation (n = 158) cohorts and followed for two years. Plaque regions of interest were manually segmented. Radiomics features were extracted with PyRadiomics and DL features from convolutional neural networks. In the training set, feature reduction used t-tests, Pearson correlation, and least absolute shrinkage and selection operator (LASSO) regression. Rad, DL, and fusion deep learning radiomics (DLR) models were developed and evaluated by receiver operating characteristic analysis. Clinical imaging predictors were screened by logistic regression and integrated with the optimal model.
Results or Findings: Hypertension, plaque ulceration, and plaque length were independent predictors (all P < 0.05). After selection, 12 Rad, 9 DL, and 16 DLR features remained. The DLR model achieved the highest AUCs in the training and validation cohorts (0.853 and 0.840) and was chosen as optimal. The combined model, which integrates DLR scores with independent clinical imaging predictors, further improved discrimination, achieving AUCs of 0.918 and 0.911 in the training and validation cohorts, respectively.
Conclusion: A combined model integrating CTA-based Rad, DL, and clinical imaging predictors accurately stratifies two-year ischemic stroke risk in asymptomatic carotid plaque patients and may support personalized risk management.
Limitations: The retrospective nature of this study may introduce inherent bias.
Funding for this study: This work was supported by the National Key Research and Development Program of China (2024YFC2417600) and the Zhejiang Medicine and Health Science and Technology Project (2025KY495, 2024KY568).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Ethical clearance was obtained from the institutional review board of the Fifth Hospital of Wenzhou Medical University.

Author Block: T. R. Readford¹, G. Martínez Rodríguez², S. Patel¹, M. Ugander³, P. Kench¹, N. Giannotti¹; ¹Sydney/AU, ²Santiago/CL, ³Solna/SE
Purpose: Dynamic contrast-enhanced MRI (DCE-MRI) can non-invasively characterise carotid atherosclerotic plaque vulnerability by quantifying measures related to arterial wall neovascularisation and endothelial permeability, which are linked to stroke risk. However, quantitative perfusion metrics such as Ktrans and Kep require complex kinetic modelling and generate disagreement regarding optimal methodology, hindering clinical interpretation.

The purpose of this study was to assess the reproducibility of a simple method for carotid plaque DCE quantification using signal intensity in the vessel wall normalised to skeletal muscle signal intensity, as an accessible surrogate marker of neovascularisation.
Methods or Background: This was a sub-study of the CAPRI trial, which had a neutral outcome for the effect of colchicine versus placebo on carotid atherosclerotic plaque volume after six months. All participants underwent T1-weighted black-blood DCE-MRI of the carotid arteries at baseline and six months, with images acquired every 10 seconds over 210 seconds following intravenous administration of 0.1 mmol/kg gadoteric acid. Plaque core and remote unaffected vessel wall were manually delineated in the same slice. Signal intensities were normalized to skeletal muscle signal intensity from the same slice and timepoint.
Results or Findings: Among included patients (n=28, median [interquartile range] age 72 [64–74] years, 36% female), normalised peak signal intensity of the plaque core was greater than the remote vessel wall at both baseline (3.5 [2.3–4.1] vs 2.1 [1.7–2.5], p<0.001) and six months (3.2 [2.5–4.4] vs 2.0 [1.7–2.5], p<0.001; but compared to baseline: p≥0.81 for both, mean±SD differences 0.7±0.7 and 0.6±0.4, respectively).
Conclusion: DCE-MRI normalised peak intensity was greater for plaque core compared to remote vessel wall, and both measurements are reproducible over six months. This simplified quantification approach may facilitate future assessment of neovascularisation and plaque phenotypes in clinical practice.
Limitations: Lack of
Funding for this study: No funding was received for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study was approved by the local human subject research ethics committee and all subjects provided written informed consent.

Author Block: G. Lin, W. Hu, W. Chen, M. Chen, J. Ji; Lishui/CN
Purpose: Our work aimed to assess the clinical value of dual-energy computed tomography (DECT) parameters in predicting in-stent restenosis (ISR) after carotid artery stenting (CAS) and to develop a nomogram model incorporating these parameters to enhance the accuracy of ISR risk prediction.
Methods or Background: Our retrospective multicenter research enrolled 205 patients who underwent CAS during January 2018 to June 2023, with DECT scans performed prior to the procedure. Two radiologists independently measured the DECT parameters and evaluated the conventional computed tomography angiography characteristics of carotid plaques. Univariate and multivariate analyses were conducted to identify independent predictors of ISR. Three models were developed: clinical model, DECT model, and nomogram model. These models were assessed based on the area under the curve (AUC) and calibration, with their clinical value assessed utilizing decision curve analysis (DCA).
Results or Findings: Among 205 patients, 35 in training set and 15 in validation set experienced ISR. Multivariate analysis identified plaque length, fat fraction, normalized iodine concentration, and effective atomic number as independent predictors of ISR. Nomogram model, combining clinical and DECT parameters, demonstrated high accuracy in predicting ISR, with 0.931 AUC in training set and 0.872 in validation set, exceeding both clinical (AUC: 0.797, 0.673) and DECT models (AUC: 0.880, 0.853). Calibration curve and DCA exhibited the nomogram’s excellent performance and clinical utility.
Conclusion: The nomogram model integrating DECT parameters and clinical predictors provides a reliable and noninvasive tool for predicting ISR risk following CAS. It facilitates formulation of individualized treatment strategies and offers important references for early intervention.
Limitations: First, the sample size was relatively small. Second, the retrospective design may have introduced selection bias. Third, manual ROI delineation may have introduced measurement errors.
Funding for this study: This work was supported by the National Key Research and Development Program of China (2024YFC2417600) and the Zhejiang Medicine and Health Science and Technology Project (2025KY495, 2024KY568).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Ethics committees of the Fifth Affiliated Hospital of Wenzhou Medical University (Center 1) and Second Affiliated Hospital of Wenzhou Medical University (Center 2) (no. 2025[I]-130-01) approved our retrospective multicenter research.

Author Block: Q. Wu, C. Huang, X. Zhu, J. Zhang, Y. Shi; Baise/CN
Purpose: To develop and validate an interpretable machine learning model that integrates clinical data with radiomics features from white matter hyperintensities (WMH) to predict haemorrhagic transformation (HT) after endovascular therapy (EVT) for acute ischaemic stroke (AIS).
Methods or Background: In this dual-centre retrospective study, we included AIS patients who underwent EVT. WMH were automatically segmented on pre-procedural T2-FLAIR MRI. Three models were developed and compared: a clinical model, a radiomics model, and a combined model. Performance was assessed using the area under the receiver operating characteristic curve (AUC). Model interpretability was investigated using SHapley Additive exPlanations (SHAP). Mediation and correlation analyses were performed to explore the relationship between key predictors and HT.
Results or Findings: The study included 450 patients, with HT occurring in 174 (38.6%). The combined model achieved the highest predictive performance in the external validation with an AUC of 0.92 (95% CI: 0.87−0.96). This was significantly superior to both clinical model (AUC: 0.78, 95% CI: 0.71−0.84) and radiomics model (AUC: 0.85, 95% CI: 0.79−0.90). SHAP analysis identified admission NIHSS score, atrial fibrillation, and several WMH textural features as the most impactful predictors. Mediation analysis revealed that WMH textural heterogeneity partially mediated the effect of chronic hypertension on the risk of HT.
Conclusion: A model combining clinical data and WMH radiomics accurately predicts post-EVT haemorrhagic transformation. SHAP revealed that textural characteristics of WMH are critical drivers of risk, providing deeper insights into the pathophysiology of vascular fragility.
Limitations: Firstly, its retrospective nature introduces a potential for selection and information bias. Secondly, although this was a dual-center study, the model's generalizability to a broader and more heterogeneous international population requires further validation. Finally, the radiomics signature's robustness may be influenced by inter-scanner variability in MRI acquisition and parameters.
Funding for this study: This study was financially supported by the National Natural Science Foundation of China (Grant No. 82460226), the Natural Science Foundation of Guangxi Autonomous Region (Grant No. 2023GXNSFAA026383, 2025GXNSFHA069156, 2025GXNSFBA069280).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Author Block: P. Reschke, K. Eichler, T. Vogl, L. D. Grünewald; Frankfurt/DE
Purpose: Cardiovascular disease is the leading cause of death in the Western world, with vascular calcification serving as a marker of advanced pathology. The Agatston score is only applicable in non-contrast CT scans. However, no standardized method exists for quantifying aortic calcification on contrast-enhanced CT scans.
Methods or Background: This retrospective study included 1125 patients (344 women, 781 men) who underwent triphasic CT angiography. Patients with metallic stents, prosthetic grafts, or poor image quality were excluded. Fixed (100–900 HU) and dynamic thresholds were applied to arterial and venous phases for aortic calcium quantification and compared with non-contrast Agatston scores.
Results or Findings: In the venous phase, the 300 HU threshold demonstrated the strongest correlation with the Agatston score (thoracic aorta Pearson’s r = 0.81,; abdominal aorta r = 0.83; all p < 0.001) and the highest predictive accuracy (R² = 0.66 and 0.68). In the arterial phase, higher thresholds were required, with optimal performance at 900 HU for the thoracic aorta (r = 0.78, R² = 0.51, p < 0.001) and 800 HU for the abdominal aorta (r = 0.80, R² = 0.45, p < 0.001). Dynamic thresholding performed significantly worse in both phases, with R² values of 0.56 (venous) and 0.07 (arterial) for thoracic calcifications and 0.62 (venous) and 0.09 (arterial) for abdominal calcifications (all p < 0.05).
Conclusion: A fixed threshold of 300 HU in the venous phase enables accurate aortic calcium quantification on contrast-enhanced CT, closely aligning with Agatston scoring and potentially allowing broader clinical application without the need for additional non-contrast imaging.
Limitations: The fixed HU thresholds used may not be universally applicable across different CT protocols, as variations in the contrast agent volume, flow rate, and contrast concentration can influence HU measurements.
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The study was approved by the ethics committee in Frankfurt.

Author Block: J. M. Brendel, I. Hadzic, T. Mayrhofer, L. H. Cooke, E. Yucel, N. K. Patel, V. Raghu, B. Ghoshhajra, B. Foldyna; Boston, MA/US
Purpose: To determine whether periaortic adipose tissue (PAAT) attenuation predicts long-term mortality in transcatheter aortic valve replacement (TAVR) patients.
Methods or Background: Higher PAAT attenuation has been linked to periaortic inflammation and can be measured on routine preprocedural CT. Its prognostic value in patients undergoing TAVR remains unclear.

In this retrospective multicenter study, we analyzed preprocedural CT scans from consecutive TAVR patients between 2014 and 2023. PAAT was segmented using a dedicated deep learning tool (TotalSegmentator) that delineates the entire aorta, followed by logical operations creating a 10-mm radial cylinder around the aortic wall with attenuation thresholds of (–190 to –30 HU). Associations with all-cause mortality were assessed using Cox regression adjusted for CT technical parameters (tube voltage, signal-to-noise ratio, slice thickness) and the Society of Thoracic Surgeons (STS) risk score.
Results or Findings: Among 928 patients from four tertiary medical centers (mean age 81±8 years; 58% male), 222 (23.9%) died during a median follow-up of 22 (14–36) months. Mean PAAT attenuation was -77.3±7.2 HU and was higher in women (+2.1±0.5 HU), nonobese patients (+4.7±0.5 HU), and those with moderate or high (≥4%) STS risk score (+1.7±0.5 HU), all p<0.001. Each 10 HU increase in PAAT attenuation was associated with an 82% higher risk of death independent of STS risk (aHR 1.82: 95%CI: 1.36–2.43; p<0.001). Patients with PAAT attenuation above -77 HU had nearly a two-fold greater mortality risk (aHR 1.97; 95%CI: 1.36–2.85, p<0.001).
Conclusion: High PAAT attenuation represents a novel opportunistic CT marker that predicts mortality in TAVR patients, and may inform long-term risk stratification and management.
Limitations: No histopathologic validation of PAAT attenuation was available, limiting mechanistic interpretation.
Funding for this study: N/A
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The institutional review board approved the study protocol, with a waiver for written informed consent.

Author Block: M. Kadıoğlu, S. Ozkök; Istanbul/TR
Purpose: To evaluate microstructural differences in periaortic adipose tissue among patients with Takayasu arteritis, atherosclerosis, and healthy controls using CT-based radiomics, and to investigate the potential of radiomic parameters as imaging biomarkers for subclinical inflammation
Methods or Background: A total of 66 subjects were included: 26 patients with Takayasu arteritis, 20 with atherosclerotic vascular disease, and 20 healthy controls. For the Takayasu group, pre-treatment CT angiograms obtained at the time of diagnosis were analyzed. All scans were performed using Philips 128-slice CT system (100–120 kVp, 1 mm slice thickness, B kernel).
The aorta was divided into ascending, aortic arch, descending, and abdominal segments. Periaortic adipose tissue was semi-automatically segmented in 3D Slicer v5.8.1, and over 300 radiomic features were extracted using PyRadiomics, following standardization guidelines proposed by Koçak B et al.
Group comparisons were conducted using ANOVA and Kruskal–Wallis tests with FDR correction (p < 0.05 considered significant).
Results or Findings: A total of 284 segments were analyzed. Statistically significant radiomic differences were detected in the ascending (n = 91), aortic arch (n = 178), descending (n = 199), and abdominal aorta (n = 189) segments.
Key discriminative parameters included Entropy, Joint Energy, Busyness, and the 90th percentile intensity, all showing p < 0.001–0.01.
Compared with atherosclerotic patients, Takayasu arteritis demonstrated markedly higher heterogeneity and entropy metrics.
Notably, even CT segments without visible wall thickening or enhancement revealed microstructural alterations in Takayasu patients.
Conclusion: CT-based radiomics can identify inflammation-related textural alterations in periaortic adipose tissue, even in morphologically normal aortic segments.
Compared with atherosclerosis, Takayasu arteritis exhibits significantly higher heterogeneity parameters, suggesting that perivascular fat radiomics may serve as a non-invasive biomarker of early vascular inflammation and tissue remodeling.
Limitations: Single-center study, limited sample size; lack of histopathologic correlation.
Funding for this study: No financial support was received.
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: P. Jessua¹, R. Sartoris¹, M. Ronot¹, L. Garzelli²; ¹Clichy/FR, ²Paris/FR
Purpose: To assess whether the mesenteric calcium score (MCS) predicts mortality or morbidity in acute occlusive arterial mesenteric ischaemia (AOAMI), of embolic or atherothrombotic origin.
Methods or Background: Retrospective cohort study of patients with AOAMI admitted to our intestinal stroke unit. Clinical, biological, imaging, treatment and outcomes data were analysed. MCS of the superior mesenteric artery was calculated using the Agatston method and compared between survivors and non-survivors at 30 days. The primary endpoint was 30-day mortality; the secondary endpoint was a composite of 30-day mortality and morbidity (short bowel syndrome, permanent stoma or home parenteral nutrition). Inter- and intra-observer reproducibility was assessed with intraclass correlation coefficients (ICC).
Results or Findings: Among 506 patients screened, 179 were included (77 women, mean age 69.5 years). Thirty-day mortality was 13% with a median overall survival of 101 months. Morbidity was significant: short bowel syndrome (27%), permanent stoma (17%), and long-term parenteral nutrition (26%), with a composite morbidity score ≥1 in 37%. Median MCS was 27, differing by etiology (embolic: 0, atherothrombotic: 375). However, no significant association was foun between MCS and 30-day mortality (p=0.34) or composite mortality-morbidity (p=0.40). Subgroup analyses by etiology confirmed these findings (embolic p=0.5 and 0.34, atherothrombotic p=0.08 and 0.63). Reproducibility of MCS were excellent (inter-observer ICC 0.94, intra-observer ICC 0.98).
Conclusion: The superior mesenteric artery calcium score does not predict short-term mortality or morbidity in AOAMI, including in atherothrombotic patients.
Limitations: Retrospective design, potential missing data, and limited power given to the relatively low mortality rate.
Funding for this study: None
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: M. Ahmad, J. Shalhoub, D. Amiras, A. G. Rockall, A. Davies; London/UK
Purpose: Diabetic foot disease (DFD) is a complex disease and is associated with lower limb amputation. Magnetic resonance imaging (MRI) is commonly used in patients with DFD. Early detection of surrogate markers on MRI may help to predict the risk of limb loss. Our aim was to explore the use of deep learning models to identify predictors on MRI for limb loss in patients with DFD.
Methods or Background: Sarcopenia severity grading and the pseudo fat fraction values calculated from 824 T1-weighted coronal MRI scans of the foot across 427 patients were combined with patient metadata to identify salient predictors for limb loss, major amputation and death and all amputation and death as composite outcomes. Least Absolute Shrinkage and Selection Operator (LASSO), Ridge Regression and Random Forest (RaF) models were used. Further analysis was also undertaken using Cox proportional hazard.
Results or Findings: The converging variables which appeared in all three models were Haemoglobin, HbA1c and renal status. The overlapping variables across ridge regression and RaF for all amputations were C-reactive protein (CRP), age, albumin and the pseudo fat fraction. Further analysis with a cox proportional hazard model found CRP was a strong predictor (Hazard Ratio 1.004 [p-value = 0.02]). Former history of transient ischaemic attacks was identified as a predictor for all amputation/death on LASSO and ridge regression. Pseudo fat fraction, weight and HbA1c were highlighted as features of importance for composite outcomes (major lower limb amputation/death and all amputation/death) on random forest.
Conclusion: There may be merit in combining MRI based image findings with patient metadata to stratify risk.
Limitations: Further internal and external validation is required with a more heterogenous cohort of patients.
Funding for this study: Imperial College Healthcare NHS Trust radiology pump priming fund
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Ethics approval from the Health Research Authority UK

Author Block: M. Ahmad, K. C. Soh, J. Shalhoub, D. Amiras, A. Davies, A. G. Rockall; London/UK
Purpose: Diabetic foot disease (DFD) is a complex disease and is associated with lower limb amputation. Magnetic resonance imaging (MRI) is commonly used in patients with DFD.
Methods or Background: 419 MRI reports were reviewed to identify salient predictors for no amputation versus major versus minor limb loss from a list of features including collections, ulcer, osteomyelitis, diabetic myopathy/sarcopenia and Charcot foot. Random Forest, Extreme Gradient Boosting and a Multilayer Perceptron model was applied. Further composite analysis was undertaken to compare no amputation versus major lower limb amputation, no amputation versus all amputation and death and all amputation versus death.
Results or Findings: Extreme gradient boosting comparing the no amputation versus minor amputation versus major amputation group using extreme gradient boosting was the only model of statistical significance. The model had a 63.4% accuracy [95% CI 0.52-0.7; p-value <0.001]. The ROC-AUC values were 0.724 (no amputation group), 0.779 (minor amputation group) and 0.439 (major amputation group). Features highlighted as potential predictors included absence of osteomyelitis and presence of Charcot as variables which influenced the outcome.
Conclusion: MRI reports may provide useful information in highlighting features which could potentially predict the risk of adverse outcomes.
Limitations: Variability in reporting can influence the outcome. Uniformity in reporting may allow this to be explored further and to create a standardised prediction model.
Funding for this study: Imperial College Healthcare NHS Trust radiology pump priming fund.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Ethics approval received from Health Research Authority UK

Author Block: J. Csőre¹, A. Crichton², E. Pomozi¹, J. Lamichhane², A. B. Lumsden², T. L. Roy²; ¹Budapest/HU, ²Houston, TX/US
Purpose: Endovascular treatment failure in chronic limb-threatening ischemia (CLTI) is not fully understood, and patient selection is crucial for optimal results. Ultrashort Echo Time (UTE) MRI is a non-contrast method shown ex vivo to characterize plaque composition. This study assessed whether MRI-defined plaque morphology influences procedural difficulty and immediate technical failure (ITF) in peripheral vascular interventions (PVI).
Methods or Background: Patients with CLTI undergoing PVI at a tertiary vascular center were enrolled and underwent 3T UTE-MRI. Lesions were classified as predominantly soft (>50% thrombus, cholesterol/lipid plaque) or hard (calcific/collagenous plaque). Operators were blinded to MRI findings and used standard pre-procedural imaging. Primary outcome was ITF; secondary outcome was lesion crossing time.
Results or Findings: A total of 43 patients (86 lesions; mean age 67.5 years, 40% female) were included; 14 lesions were excluded (7 diagnostic DSA only, 7 non-diagnostic MRI). Of the 72 evaluable lesions, 66.7% were scored as soft and 33.3% were hard. ITF occurred in 15.3% of lesions, most commonly due to crossing failure (72.7%). Hard lesions had a markedly higher ITF rate compared with soft (41.7% vs 2.1%, p<0.001, Chi²=19.3). In a subgroup analysis among >75% stenosis/occlusions, ITF was 58.8% in hard vs 4.3% in soft lesions (p<0.001, Chi²=14.5). Median crossing time was significantly prolonged in hard lesions (535s vs 38s, p=0.024).
Conclusion: UTE-MRI enables in vivo characterization of vessel wall composition and predicts procedural complexity and ITF in PVI. Despite its potential, assessment of plaque morphology remains underused in CLTI care. Incorporating UTE-MRI into clinical workflows may refine patient selection, guide interventional strategy, and improve outcomes.
Limitations: Single-center study, sample size
Funding for this study: Jerold B. Katz Academy of Translational Science under project number 15790002 (recipient's name: Trisha Roy), the American Heart Association Transformational Award under project ID 17590004 (recipient's name: Trisha Roy), and the National Institutes of Health Research Project grant (R01) under award number R01HL174587 (recipient's name: Trisha Roy).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Study ID 15790002