Federal Project Grant Award Summary

New York University School of Medicine received a $700,171 Project Grant award dated July 16, 2026, from the National Cancer Institute under the Cancer Detection and Diagnosis Research program (CFDA 93.394) to advance clinically translatable deep learning (DL) models for prostate cancer detection on magnetic resonance imaging (MRI). The project, with a completion date of June 30, 2031, addresses critical gaps in prostate MRI interpretation by developing robust artificial intelligence solutions that improve diagnostic accuracy, consistency, and clinical reliability. The work encompasses three primary objectives: training a novel DL model on diverse imaging datasets to ensure consistent performance across varied scanners, protocols, and patient populations; incorporating uncertainty quantification and out-of-distribution detection mechanisms to enhance prediction reliability and provide actionable confidence measures for clinical decision-making; and evaluating the clinical utility of artificial intelligence assistance and confidence visualization through a multi-reader study.

This research directly supports the National Cancer Institute's strategic focus on improving screening, early detection, and diagnostic techniques for cancer by reducing false positives and false negatives in prostate cancer detection, thereby decreasing unnecessary biopsies and improving patient management and surveillance outcomes.