Same-Day Diagnostic Imaging in Breast Cancer Screening

Breast cancer screening mammography has been shown to decrease mortality, but it is limited by inefficiencies such as false-positive results (approximately 95% of recalled exams are false positives) and significant wait times for further diagnostic evaluation, which cause considerable patient anxiety. The overarching goal of this project is to minimize practice variabilities associated with recalls, reduce patient anxiety, and increase patient satisfaction.

The project proposes an online interpretation, same-day diagnostic imaging paradigm by introducing an artificial intelligence (AI) solution at the point of care to triage screening exams. This approach aims to reduce the overall callback rate, increase patient satisfaction by providing immediate screening results and, for women requiring further diagnostic workup, eliminate the delay between screening and diagnostic workup by enabling real-time interpretation and same-day scheduling. The project is structured around three aims:

• Validate and integrate an AI algorithm to triage screening mammograms within the institution's breast screening population, ensuring non-inferior performance to radiologists and clear communication of results.
• Design and assess an AI-enabled workflow for same-day diagnostic exams, analyzing current care and identifying impediments.
• Implement and evaluate the impacts of an AI-enabled same-day diagnostic imaging paradigm in three stages (pilot, implementation, expansion) across multiple imaging centers.

UCLA Health, with its large volume of screening exams (>40,000 annually), provides a unique environment for evaluating this paradigm. The expected outcome is a generalizable approach for evaluating and integrating AI algorithms to improve healthcare delivery.

Contact PI: William Hsu

Funding Source: DHHS AHRQ