Our research within the domain of clinical decision support systems encompasses the integration of machine learning models with various medical imaging modalities. This integration aims to improve the accuracy and efficiency of medical diagnostics and prognostics. Specifically, we focus on the segmentation of areas of interest (AOIs) in imaging data, direct extraction of features from images or segmented regions, estimation of disease progression, and the utilization of multimodal approaches to refine predictive accuracy. These methodologies not only facilitate the diagnosis and monitoring of diseases but also assist in the identification of biomarkers, providing significant support to clinicians in decision-making processes.