A New Fast Framework for Early Detection of Prostate Cancer Without Prostate Segmentation

Abstract

This paper presents a computer-aided diagnosis (CAD) system for early detection of prostate cancer from diffusion-weighted magnetic resonance imaging (DWI) acquired at six different b-values. Our system starts by defining a region of interest (ROI) that includes the prostate across the different slices of the input DWI volume. Then, the apparent diffusion coefficient (ADC) of the defined ROI is calculated, normalized and refined. Then, the probability density functions (PDFs) of the refined ADC volumes at the distinct b-values are constructed. Finally, the classification of prostate into either benign or malignant is achieved using a classification system of two stages. The proposed system is the first system of its type that has the ability to detect prostate cancer without any prior processing (e.g., the segmentation of the prostate region). Evaluation of the proposed system is done using DWI datasets acquired from 45 patients (20 benign and 25 malignant) at six distinct b-values. The acquisition of these DWI datasets is performed using two different scanners with distinct magnetic field strengths (1.5 Tesla and 3 Tesla). The resulting area under curve (AUC) is 0.77, which shows that the proposed system approaches the state-of-the-art performance without any prior processing.