A Comparative Study of Automated Deep Learning Segmentation Models for Prostate MRI

SIMPLE SUMMARY: Prostate cancer represents a highly prevalent form of cancer worldwide, with timely detection and treatment being crucial for achieving a high survival rate. Manual segmentation, which is the process of manually identifying different anatomical structures or tissues within an image, is the most prevalent detection method. However, it is a time-consuming and subjective task constrained by the radiologists’ expertise, which underpins the demand for automated segmentation methods. In this study, we conduct a comprehensive and rigorous comparison of multiple prevalent deep learning-based automatic segmentation models for the prostate gland and both peripheral and transition zones, using multi-parametric MRI data. ABSTRACT: Prostate cancer is one of the most common forms of cancer globally, affecting roughly one in every eight men according to the American Cancer Society. Although the survival rate for prostate cancer is significantly high given the very high incidence rate, there is an urgent need to improve and develop new clinical aid systems to help detect and treat prostate cancer in a timely manner. In this retrospective study, our contributions are twofold: First, we perform a comparative unified study of different commonly used segmentation models for prostate gland and zone (peripheral and transition) segmentation. Second, we present and evaluate an additional research question regarding the effectiveness of using an object detector as a pre-processing step to aid in the segmentation process. We perform a thorough evaluation of the deep learning models on two public datasets, where one is used for cross-validation and the other as an external test set. Overall, the results reveal that the choice of model is relatively inconsequential, as the majority produce non-significantly different scores, apart from nnU-Net which consistently outperforms others, and that the models trained on data cropped by the object detector often generalize better, despite performing worse during cross-validation.