Prostate Segmentation in Magnetic Resonance Images Using Artificial Neural Networks: A Systematic Literature Review

Currently, Artificial Neural Networks have become a very popular machine learning approach for the automatic segmentation of medical images. To understand how the field of automatic prostate segmentation using artificial neural networks over MRIs has evolved lately, we present a Systematic Literature Review (SLR) on this topic. This SLR is based on the Kitchenham methodology applied to 372 primary studies published from 2015 to 2020. Inclusion and exclusion criteria were applied to reduce the number of papers to 65. The contributions of this SLR consist of determining the segmentation structure, type of segmentation, standard architectures, and the metrics used. We also identified findings that could open new directions of work for future research.