Abstract #4963

Spatially-sensitive model for detection of prostate cancer on multiparametric MRI

Ethan Leng¹, Jin Jin², Lin Zhang², Christopher A. Warlick³, Benjamin Spilseth⁴, Joseph S. Koopmeiners², and Gregory J. Metzger¹

¹Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, MN, United States, ²Division of Biostatistics, School of Public Health, University of Minnesota, Minneapolis, MN, ³Department of Urologic Surgery, Institute of Prostate and Urologic Cancers, University of Minnesota, Minneapolis, MN, United States, ⁴Department of Radiology, University of Minnesota, Minneapolis, MN, United States

A novel predictive model of prostate cancer (PCa) on multiparametric MRI was developed that takes into account the spatial distribution of PCa within the prostate and the spatially-autocorrelated nature of mpMRI data. The performance of the proposed model was compared to the LASSO-based model we previously described on 34 PCa cases using both voxel-wise metrics (AUC) and slice-wise metrics ($$$s_s$$$) we recently developed. The proposed model achieved superior predictive performance both in terms of AUC (0.81 vs 0.77) and $$$s_s$$$ (0.45 vs. 0.35) over the 34 cases, with significant improvements for the majority of cases.

This abstract and the presentation materials are available to members only; a login is required.

Join Here