Abstract:Journal of Endovascular Therapy, Ahead of Print. Objectives:Our study aimed to investigate the correlations between radiomics-based assessment and outcomes, including positive aortic remodeling (PAR), reintervention for dissection at 1 year, and overall survival, in patients with Type B aortic dissection (TBAD) who underwent thoracic endovascular aortic repair (TEVAR).Methods:This was a single-center, retrospective, cohort study. The cohort comprised 104 patients who had undergone TEVAR of TBAD in our institution between January 2010 and October 2022. We segmented preoperative computed tomography (CT) images of the patients' descending aorta regions, then extracted a comprehensive set of radiomic features, including first-order features, shape features (2D and 3D), gray-level co-occurrence matrix (GLCM), gray-level size zone matrix, gray-level run length matrix, gray-level dependence matrix, neighborhood gray-tone difference matrix, from the regions of interest. Next, we selected radiomics features associated with total descending aorta positive aortic remodeling (TDA-PAR) and reintervention by least absolute shrinkage and selection operator (LASSO) regression and features associated with survival by LASSO-Cox regression. This enabled us to calculate radiomics-based risk scores for each patient. We then allocated the patients to high and low radiomics-based risk groups, the cutoff being the median score. We used 3 different models to validate the radiomics-based risk scores.Results:The patients' baseline characteristics did not differ between those who achieved TDA-PAR and those who did not. The radiomics-based risk scores were significantly and independently associated with all 3 outcomes. As to the impact of specific radiomics features, we found that GLSZM_SmallAreaLowGrayLevelEmphasis and shape_Maximum2DDiameterColumn had positive impacts on both reintervention and survival outcomes, whereas GLCM_Idmn positively affected survival but negatively affected reintervention. We found that radiomics-based risk for TDA-PAR correlated most significantly with zone 6 PAR.Conclusions:Radiomics-based risk scores were significantly associated with the outcomes of TDA-PAR, reintervention, and overall survival. Radiomics has the potential to make significant contributions to prediction of outcomes in patients with TBAD undergoing TEVAR.Clinical ImpactIn this study of 104 patients with Type B aortic dissection, we demonstrated associations between radiomics-based risk and postoperative outcomes, including total descending aorta positive aortic remodeling, reintervention and survival. These findings highlight radiomics' potential as a tool for risk stratification and prognostication in acute Type B aortic dissection management.

Unsupervised machine learning cluster analysis to identification EVAR patients clinical phenotypes based on radiomics

A Radiomics Model for Predicting the Outcome of Endovascular Abdominal Aortic Aneurysm Repair Based on Machine Learning

Development and Comparison of Multimodal Models for Preoperative Prediction of Outcomes After Endovascular Aneurysm Repair

Identification of Clinical Heterogeneity and Construction of Prediction Models for Novel Subtypes in Patients with Abdominal Aortic Aneurysm: An Unsupervised Machine Learning Study

Radiomics and machine learning to predict aggressive type 2 endoleaks after endovascular aneurysm repair: a proof of concept

A radiomics model based on aortic computed tomography angiography: the impact on predicting the prognosis of patients with aortic intramural hematoma (IMH)

Deep learning-based radiomics of computed tomography angiography to predict adverse events after initial endovascular repair for acute uncomplicated Stanford type B aortic dissection

Assessing abdominal aortic aneurysm growth using radiomic features of perivascular adipose tissue after endovascular repair

Unraveling Phenotypic Heterogeneity in Stanford Type B Aortic Dissection Patients Through Machine Learning Clustering Analysis of Cardiovascular CT Imaging

Radiogenomics for Predicting Microsatellite Instability Status and PD-L1 Expression with Machine Learning in Endometrial Cancers: A Multicenter Study

Computed tomography-based radiomics model to predict adverse clinical outcomes in acute pulmonary embolism

CT Angiography Radiomics Combining Traditional Risk Factors to Predict Brain Arteriovenous Malformation Rupture: a Machine Learning, Multicenter Study

A DWI-based radiomics-clinical machine learning model to preoperatively predict the futile recanalization after endovascular treatment of acute basilar artery occlusion patients

Radiomics-Based Risk Assessment Correlates With Outcomes in Patients With Acute Type B Aortic Dissection Undergoing Thoracic Endovascular Repair

Construction of prediction models for novel subtypes in patients with arteriosclerosis obliterans undergoing endovascular therapy: an unsupervised machine learning study

Cardiac ultrasomics for acute myocardial infarction risk stratification and prediction of all-cause mortality: a feasibility study

Prediction of the Endoleak Risk after EVAR Based on Multimodal Geometric Characteristics Before AAA Surgery

Construction of Prediction Model Integrating the Radiomics Features of Segmented Clots for Risk Stratification in Acute Pulmonary Embolism

CT-based radiomics: predicting early outcomes after percutaneous transluminal renal angioplasty in patients with severe atherosclerotic renal artery stenosis

CT texture analysis predicts abdominal aortic aneurysm post-endovascular aortic aneurysm repair progression

The radiomics-clinical nomogram for predicting the response to initial superselective arterial embolization in renal angiomyolipoma, a preliminary study