Project Grant R01CA308887

Federal Grant Award Summary

The National Cancer Institute (NCI) awarded The Leland Stanford Junior University a $660,346 Project Grant effective July 1, 2026, through June 30, 2031, under the Cancer Detection and Diagnosis Research program (CFDA 93.394). The grant supports development of artificial intelligence (AI) models to enable real-time prostate cancer detection during ultrasound-only biopsies (PROCUS). The deliverables include multi-vendor AI models trained on brightness (B)-mode conventional and micro-ultrasound data from over 5,000 exams across five ultrasound vendors to ensure generalization and clinical workflow integration. The research addresses a significant clinical gap: while magnetic resonance imaging (MRI)-guided targeting improves cancer detection by 40%, it remains unavailable or cost-prohibitive in approximately 64% of biopsies performed annually in the United States.

The project leverages Stanford's unique multi-vendor cohort with biopsy-confirmed lesions and whole-mount pathology labels from 110 men who underwent radical prostatectomy. Preliminary studies demonstrated that supervised AI models trained on single-vendor conventional ultrasound data detected clinically significant prostate cancers at performance levels equivalent to radiologists interpreting MRI. The completed project is expected to transform ultrasound from a passive imaging tool to an active targeting modality, potentially improving prostate cancer diagnosis in over 55,000 additional men annually by enabling clinicians to identify and target suspicious lesions in ultrasound-only settings where MRI guidance is unavailable.