2024 World Conference on Lung Cancer (WCLC) - Posters-P4.04C.05 AI-Assisted CXR Analysis in the Detection of Lung Nodules and Incidental Lung Cancers

Pdf Summary

The study conducted by Se Hyun Kwak and colleagues at Yonsei University College of Medicine explores the impact of AI in interpreting chest X-rays (CXRs) to identify lung nodules as potential indicators of lung cancer and other conditions. Investigating data from March 2021 to January 2023 at Yongin Severance Hospital, the study utilized the AI-based software Lunit INSIGHT, integrated into clinical workflows to detect various lung lesions. <br /><br />The researchers found that AI-detected nodule scores were notably higher in patients who underwent advanced diagnostic procedures such as chest CT scans, proving more accurate in cases confirmed as lung cancer compared to false positives. Additionally, lung cancer patients exhibited higher mean nodule scores compared to those with active pulmonary infections or old inflammatory sequelae. The outcomes demonstrate AI's effective role in distinguishing between malignant and non-malignant lung conditions.<br /><br />Importantly, AI's integration into CXRs led to significant early detection of potentially serious conditions, compelling clinical actions like specialist consultations and further diagnostics. Among the 36 lung cancer cases identified in the study, adenocarcinoma was the most common subtype, followed by squamous cell carcinoma and small cell lung cancer, with diverse treatment strategies applied including surgery and chemotherapy.<br /><br />The study concludes that AI not only aids in early detection of lung abnormalities but also underscores the necessity for integrating efficient alerts and follow-up strategies into AI systems to enhance diagnostic accuracy and patient management. As such, AI demonstrates promising potential in reshaping diagnostic and treatment pathways for lung-related conditions in outpatient settings.

Asset Subtitle

Se Hyun Kwak

Meta Tag

Speaker Se Hyun Kwak

Topic Screening & Early Detection

Keywords

AI

chest X-rays

lung nodules

lung cancer

Lunit INSIGHT

diagnostic accuracy

adenocarcinoma

early detection

clinical workflows

Yonsei University