The goal of this project was to evaluate the feasibility of machine learning methods based on feature extraction and classification for differentiating benign and malignant prostate cancer tumors in mp-MRI with the intent of building an algorithm that best reduces over-biopsy and enables in-silico biopsy. The specific aim of this project was to evaluate various extracted texture features, dimensionality reduction, and machine learning classification methods to determine the computer-aided analysis model that provides the greatest classification accuracy for malignant and benign prostate cancer images.
