Federal Project Grant Award Summary

The National Cancer Institute awarded a $249,000 Project Grant to The Johns Hopkins University under the Cancer Detection and Diagnosis Research Program (CFDA 93.394) for a three-year project period from January 1, 2026 through December 31, 2028. The project, titled "Large-Scale Artificial Intelligence Using Radiomics, Deep Learning, and Physics-Based Generative Modeling to Improve Clinical Outcomes in Nuclear Medicine," addresses critical challenges in applying artificial intelligence and deep learning methodologies to cancer detection and diagnosis within nuclear medicine imaging.

The research deliverables include three primary components: (1) development of a large-scale clinical database of Positron Emission Tomography/Computed Tomography (PET/CT) images with physician-annotated ground truth to support machine learning model training; (2) creation of a physics-guided deep generative modeling approach to generate realistic simulated PET/CT data with known ground truth, thereby expanding training datasets without the constraints of limited annotated clinical data; and (3) quantification of radiomic feature robustness to address reproducibility challenges stemming from scanner variability, reconstruction methods, and operator-dependent segmentation differences. These integrated deliverables are designed to advance reproducible, generalizable AI-based diagnostic tools that improve early cancer detection and clinical outcomes through enhanced analysis of nuclear medicine imaging data.