iMagen - Based Analysis of the Evolution of Treated Abdominal Aortic Aneurysms by Endovascular Repair

During the second year of the MEASURE project, predictive models have been developed to assess the progression of patients treated with EVAR. These models are based on features calculated from the segmentations of thrombus or aneurysm in the patients' postoperative CTA scans. Some of these features, such as shape parameters, are commonly used, but new biomarkers based on observed deformations have also been developed. These biomarkers have proven to be highly predictive of patient outcomes. Regarding more common biomarkers, texture features extracted from the first postoperative scan are sufficient for an SVM classification model to differentiate between cases with favorable or unfavorable outcomes with 92% accuracy. These results have generated significant international scientific impact, resulting in several high-impact journal publications and conference presentations, highlighting the contributions of the agents integrated within the Basque Science, Technology, and Innovation Network.