Michael Calhoun

Ph.D.

Precision Medical Ventures, Inc.

CEO

NLST (Learn more about this study)

NLST-800

Jun 14, 2021

Development of diagnostic support software to aid clinical decision-making for lung disease

Lung nodules are a frequent finding on chest CT. Because of the difficulty in identifying cancer risk, diagnosis of lung nodules often requires expensive, time-consuming and sometimes invasive procedures. Prospective, population-based studies such as the NLST are rare resources to understand disease risk factors, and have incredible potential value for improving diagnostic precision. In addition to structured reporting and long-term follow-up, the breadth of the collected data (e.g., in factors such as geography, population and acquisition) may improve analytic explainability and provide clinical context.

This project proposes to capitalize on these resources, together with advances in machine learning and other capabilities of the applying institution. The goal is to improve lung nodule management and potentially lung disease more broadly. In addition to research goals, the project targets real-world clinical improvements through productization of robust research results.

Research associated with this project will be disseminated to the community through publication. These efforts will also be evaluated for clinical usability within products, incorporated for distribution to practitioners, and submitted for review to clinical care committees and specialty organizations.

1) Build and validate predictive models for lung nodule detection, and prediction of malignancy risk and treatment-relevant parameters.

2) Improve clinical decision-making by providing quantitative and qualitative results from trained models based on similar patients.

3) Deliver products that expand and optimize existing guidelines for lung nodule management.