This is a Project Grant awarded by the National Cancer Institute (NCI) under the Cancer Detection and Diagnosis Research program (CFDA 93.394). The $572,490 grant will fund the development of PILLAR, an AI-based tool to predict breast cancer risk from longitudinal multi-modal breast imaging data.
The key products and services to be delivered under this grant include:
Development of novel machine learning architectures and self-supervised learning algorithms to enable PILLAR to accurately predict breast cancer risk by analyzing longitudinal mammograms, tomosynthesis, and MRI data.
Creation of algorithms to improve the robustness of PILLAR's image-based AI models against unseen imaging protocols, allowing the tool to adapt and provide accurate risk assessments.
Development of a simulation framework to benchmark the performance of PILLAR and other image-based AI models in improving breast cancer screening and prevention guidelines compared to existing clinical risk models.
The grant will leverage large imaging datasets from academic medical centers and registries, and the research findings will be validated using external datasets. If successful, this project aims to yield a new class of highly accurate, AI-powered breast cancer risk assessment tools that can enhance early detection and prevention while reducing overtreatment.