Enhancing Prostate Cancer Detection with AI-Based Ensemble Models on T2-Weighted MRI Sequences

Prostate cancer poses a significant threat to men’s health, and timely detection is crucial for effective treatment [1]. Advances in artificial intelligence (AI) offer promising improvements in prostate cancer detection, potentially enhancing early detection rates, refining treatment strategies, and improving patient outcomes [2]. This study employed 5-fold cross-validation to train three models: an Attention-Res-U-Net, a Vanilla-Net, and a V-Net. Based on validation results, the best-performing average ensemble model was selected and tested using Intersection over Union (IoU) and Dice Similarity Coefficient (DSC) metrics.