Can computer-aided diagnosis assist in the identification of prostate cancer on prostate MRI? a multi-center, multi-reader investigation

The cover for issue 73 of Oncotarget features Figure 2, "Benefit of CAD in TZ tumor identification," by Gaur, et al.

For prostate cancer detection on prostate multiparametric MRI, the Prostate Imaging-Reporting and Data System version 2 and computer-aided diagnosis systems aim to widely improve standardization across radiologists and centers.

Dr. Baris Turkbey from the Molecular Imaging Program, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA said "Men with suspected or known prostate cancer are increasingly evaluated with prostate multiparametric MRI because it aids in the detection of clinically significant disease"

To address some of these issues the Prostate Imaging-Reporting and Data System version 2 was introduced in 2015 as a set of guidelines outlining standard acquisition parameters and a categorization system for cancer detection.

Figure 2: Benefit of CAD in TZ tumor identification. CAD (top left) picked up a tumor (arrows) in the right apical anterior TZ, identified by more readers on MRI (T2W top right, ADC map bottom left, b-1500 bottom right) with CAD assistance. ND = not detected, D = detected; the tumor was found by 5 readers with CAD assistance versus 1 reader with mpMRI alone. Radical prostatectomy histopathology mapping revealed Gleason 4+5 prostatic adenocarcinoma within this lesion.

Computer-aided diagnosis systems have shown promise in the identification of prostate cancer on mp MRI in several single institution studies.

The purpose of this study was to test a new prostate CAD on a highly heterogeneous, real-world, data set from 5 institutions against mp MRI interpretations with PI-RADSv2 using a diverse set of readers, varying in location and experience.

The Baris Turkbey research team concluded, "when using PI-RADSv2 and a CAD system based on heterogeneous imaging acquisitions, readers with different experience levels were able to detect index lesions with comparable sensitivity to non-assisted interpretations."

Full text - https://doi.org/10.18632/oncotarget.26100

Correspondence to - Baris Turkbey - ismail.turkbey@nih.gov

Keywords - computer-aided diagnosis, prostate cancer, multiparametric MRI, PI-RADSv2, tumor detection