Leveraging electronic medical records for early lung cancer diagnosis: An evaluation of the C the Signs AI cancer prediction platform using the Mayo data platform.

8053 Background: In the US, only 27.4% of lung cancer cases are diagnosed early, with 5-year survival rates of 63% for localized and 27% for late-stage cancers. Despite recommendations since 2012 for screening high-risk individuals with low-dose CT, uptake has been limited, and most lung cancer diagnoses occur after symptoms appear. Similarly, chest x-rays, while commonly used as an initial test to investigate patients with symptomatic suspected lung cancer, have demonstrated limited sensitivities between 50-70% and specificities over 80%. Symptoms often overlap with common conditions, making detection challenging, and studies have identified median delays of 187 days from symptom onset to diagnosis. This prolonged interval presents an opportunity for improvement. This study examines the use of the AI cancer prediction platform, C the Signs, to passively screen for lung cancer by leveraging electronic medical records (EMRs) for early lung cancer detection. Methods: Utilizing the Mayo data platform, we conducted a retrospective analysis of EMR data from 894,409 patients, including 7,395 individuals diagnosed with lung cancer. We assessed the sensitivity and specificity of the AI cancer prediction platform, in identifying patients at risk of lung cancer. Additionally, we compared the timing of lung cancer risk identification by the AI cancer prediction platform with the timing of diagnoses made by physicians to determine whether the platform enabled earlier detection. Results: The AI cancer prediction platform detected 6,749 cases of lung cancer among the 7,395 individuals diagnosed, resulting in an early detection sensitivity of 91.5%. The platform identified 423,249 false positives among the 887,014 patients who did not have lung cancer, leading to a specificity of 52.3%. Additionally, it identified the risk of a lung cancer diagnosis in 26.6% of patients up to five years earlier than the diagnoses made by physicians. Conclusions: This study highlights the potential of leveraging EMR data and AI platforms like C the Signs to enhance early lung cancer detection. Chest X-rays, with their reduced sensitivity for early-stage lesions and reliance on symptom-driven use, remain limited as a screening tool. In contrast, the AI platform achieved a sensitivity of 91.5% and identified 26.6% of cases up to five years earlier than traditional diagnoses. These findings underscore the promise of AI-based platforms as supplementary tools for improving early detection and facilitating timely intervention to enhance patient outcomes.