Skip to main content

A Computer-Aided Detection System for Prostate Cancer using Multimodal MR Imaging

Thursday, October 11, 2012 — Poster Session III

10:00 a.m. – Noon

Natcher Conference Center, Building 45



* FARE Award Winner


  • P Liu
  • S Wang
  • B Turkbey
  • P Pinto
  • P Choyke
  • R Summers


A computer-aided detection and classification system for prostate cancer (CaP) is proposed. The system features the ability to generate a prediction map of the prostate for biopsy planning. Also, the system is able to incorporate multiple MRI modalities. The performance of the proposed system was tested on a dataset of 26 CaP patients with T2-weighted and Diffusion-weighted MRI imaging (15 training and 11 testing cases). A significant improvement in area under the receiver operating curve (AUROC) was obtained (p < 0.0001) when both imaging modalities were utilized (AUROC = 85.4%) compared with when only the T2-weighted images were used (AUROC = 82.6%). Additionally, the generated prediction maps show promise in aiding clinicians with tumor detection.

back to top