Incidental Pulmonary Nodule (IPN) Programs Working Together with Lung Cancer Screening and Artificial Intelligence to Increase Lung Cancer Detection

Abstract

Current lung cancer screening guidelines in the United States fail to identify many individuals at risk of developing the disease. Additionally, existing healthcare infrastructure has been leveraged to establish IPN clinics, a promising approach to addressing the limitations of current screening guidelines. Early-stage lung cancer is frequently diagnosed because of the incidental detection of pulmonary nodules on clinically indicated chest CT scans, particularly in the absence of formal screening programs. While artificial intelligence (AI) systems for lung cancer detection have demonstrated significant advancements in medicine, their clinical validation in screening settings remains limited. This review will discuss the pivotal trials underpinning the United States Preventive Services Task Force (USPSTF) recommendations for lung cancer screening, which have shaped the current guidelines for at-risk populations. We will explore recent studies investigating the role of AI in enhancing lung cancer screening efforts, highlighting how AI has the potential to improve early detection, streamline workflows, and reduce false positives and negatives in screening processes. This review will present the lung cancer screening rates at our institution, with a specific focus on the validation and integration of AI-driven technologies into our established screening programs. Using AI algorithms, we have validated enhanced detection capabilities through retrospective analysis of historical patient data, demonstrating significant improvements in identifying high-risk individuals and early-stage malignancies. These AI models, validated through rigorous cross-validation methods and clinical trials, have proven to outperform traditional screening approaches in sensitivity and specificity. The integration of these AI technologies within the lung cancer screening framework not only optimizes existing programs but also expands access to screening, improving early detection rates and ultimately leading to better patient outcomes. Through continuous validation and refinement, we aim to solidify AI's role in transforming lung cancer detection and patient care. Through ongoing validation and implementation, AI can play a crucial role in transforming lung cancer screening practices, ultimately contributing to earlier diagnosis and improved patient survival.

Keywords: IPN clinic; artificial intelligence; incidental pulmonary nodule; lung cancer detection; lung cancer screening.

Figure 1

Memorial Cancer Institute Incidental Pulmonary Nodule Program. (a) On the left, the figure depicts the data from our traditional lung cancer screening program between July 2021 and July 2024 prior to the use of our combined IPN and AI model. (b) On the right, the figure depicts the data from our combined IPN and AI model between 26 February 2023 and 4 November 2024. There is a marked increase in lung cancer detection in the IPN and AI model from 10 to 107 diagnoses after only approximately 20 months of implementation.