ABSTRACT
Prostate cancer is the second most common cancer among men in the United States after skin cancer. Although it can be a serious disease, early diagnosis of prostate cancer can significantly prevent the growth of cancerous cells. The feature extraction is the process of defining and deriving from the prostate region computational entities that form a sort of prostate cancer signature. Full computer-aided diagnosis (CAD) systems presented in several studies have reported the use of engineered features obtained from multimodal magnetic resonance imaging (MRI) to detect prostate cancer. Similar to other medical imaging CAD systems, the computer-aided diagnosis of prostate cancer using MRI framework encompasses four stages, namely: pre-processing, prostate region extraction, features extraction, and classification. Identifying the region of interest in the MR images is essential to reduce the complexity of the next stages and enhance the performance of the overall CAD system.
