Pd50-11 Population Specific Radiomics Model Improves Prostate Cancer Diagnostic Risk Stratification at Mri in African American Men

You have accessJournal of UrologyProstate Cancer: Detection & Screening V (PD50)1 May 2024PD50-11 POPULATION SPECIFIC RADIOMICS MODEL IMPROVES PROSTATE CANCER DIAGNOSTIC RISK STRATIFICATION AT MRI IN AFRICAN AMERICAN MEN Abhishek Midya, Siddharth Balakrishnan, Sree Harsha Tirumani, Leonardo K. Bittencourt, Amogh Hiremath, Lee Ponsky, Anant Madabhushi, and Rakesh Shiradkar Abhishek MidyaAbhishek Midya , Siddharth BalakrishnanSiddharth Balakrishnan , Sree Harsha TirumaniSree Harsha Tirumani , Leonardo K. BittencourtLeonardo K. Bittencourt , Amogh HiremathAmogh Hiremath , Lee PonskyLee Ponsky , Anant MadabhushiAnant Madabhushi , and Rakesh ShiradkarRakesh Shiradkar View All Author Informationhttps://doi.org/10.1097/01.JU.0001008620.35181.96.11AboutPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookLinked InTwitterEmail Abstract INTRODUCTION AND OBJECTIVE: Radiomics and machine learning models have demonstrated improved detection of clinically significant PCa (csPCa) on Magnetic Resonance Imaging (MRI)2-5. However, they have not explicitly accounted for disparities between African American (AA) and Caucasian American (CA) men.1 We hypothesize that radiomics can be used to quantify population specific PCa heterogeneity on MRI and build risk stratification models tailored for AA men potentially allowing for improved diagnosis and treatment outcomes. METHODS: This IRB approved, HIPAA complaint retrospective study comprised of N=111 PCa patients (55 AA, 56 CA) who underwent 3Tesla MRI prior to systematic/MRI guided biopsy. Gleason grade group (GGG) from biopsy was used to define csPCa (GGG>1) and clinically insignificant PCa (ciPCa; GGG=1). 2 board certified radiologists annotated PCa regions on MRI. Radiomic features quantify texture and shape-based heterogeneity were derived from T2-weighted (T2W) MRI within PCa region9. Dataset was split into training (DT; N=85 (42 AA, 43 CA)) and validation (DV; N=26 (13 AA, 13 CA)). Random forest machine learning models were trained in conjunction with mRMR feature selection method on DT to distinguish csPCa and ciPCa for AA specific (CAA), CA specific (CCA) and race agnostic (CRA), and validated on DV. Wicoxon rank-sum test was used to assess significant differences (p<0.05) in radiomics and classification models were assessed using area under receiver operating characteristics curve (AUC). RESULTS: Radiomic 1st order statistics from T2W quantifying voxel intensities were associated with csPCa in AA men. Intensity cooccurrence radiomic features quantifying neighborhood intensity relationships on T2W were associated with csPCa among CA men. Radiomics associated with csPCa among AA men did not show significant differences in CA men and vice-versa. Population specific models CAA and CCA resulted in AUC=0.75 and 0.73 on AA and CA men respectively, while CRA resulted in AUC=0.54 and 0.55 for AA and CA patients in DV. CONCLUSIONS: Population specific radiomics model resulted in improved diagnostic risk stratification among AA men compared to a population agnostic approach. Future work involves validation on larger, multi-institutional datasets and correlation against genomic biomarkers. Download PPT Source of Funding: Research reported in this publication was supported by the National Cancer Institute under award numbers R01CA268287A1, U01CA269181, R01CA26820701A1, R01CA249992-01A1, R01CA202752-01A1, R01CA208236-01A1, R01CA216579-01A1, R01CA220581-01A1, R01CA257612-01A1, 1U01CA239055-01, 1U01CA248226-01, 1U54CA254566-01, National Heart, Lung and Blood Institute 1R01HL15127701A1, R01HL15807101A1, National Institute of Biomedical Imaging and Bioengineering 1R43EB028736-01, VA Merit Review Award IBX004121A from the United States Department of Veterans Affairs Biomedical Laboratory Research and Development Service, the Office of the Assistant Secretary of Defense for Health Affairs, through the Breast Cancer Research Program (W81XWH-19-1-0668), the Prostate Cancer Research Program (W81XWH-20-1-0851), the Lung Cancer Research Program (W81XWH-18-1-0440 and W81XWH-20-1-0595), the Peer Reviewed Cancer Research Program (W81XWH-18-1-0404, W81XWH-21-1-0345, and W81XWH-21-1-0160), (W81XWH-22-1-0236) the Kidney Precision Medicine Project (KPMP) Glue Grant, sponsored research agreements from Bristol Myers-Squibb, Boehringer-Ingelheim, Eli-Lilly, and AstraZeneca, American Cancer Society Institutional Research Grant from Winship Cancer Institute, and Winship Invest$ Pilot Grant from the Winship Cancer Institute © 2024 by American Urological Association Education and Research, Inc.FiguresReferencesRelatedDetails Volume 211Issue 5SMay 2024Page: e1061 Advertisement Copyright & Permissions© 2024 by American Urological Association Education and Research, Inc.Metrics Author Information Abhishek Midya More articles by this author Siddharth Balakrishnan More articles by this author Sree Harsha Tirumani More articles by this author Leonardo K. Bittencourt More articles by this author Amogh Hiremath More articles by this author Lee Ponsky More articles by this author Anant Madabhushi More articles by this author Rakesh Shiradkar More articles by this author Expand All Advertisement PDF downloadLoading ...