Drew Moghanaki, professor and chief of thoracic oncology in the Department of Radiation Oncology at UCLA and chief medical officer of Respirati, highlighted a University of Toronto study on LinkedIn that links artificial intelligence screening for interstitial lung disease (ILD) on chest CT to the risk of severe radiation pneumonitis in patients with locally advanced non-small cell lung cancer (NSCLC).
Researchers used a convolutional neural network to identify ILD on chest CT scans and assess risk of grade 3 radiation pneumonitis. The AI model alone modestly stratified risk, identifying roughly a 10% difference in outcomes. When AI findings were combined with expert review by a thoracic radiologist, the predicted risk rose to about 30%.
The study’s authors and Moghanaki note that, if validated, AI-augmented risk modeling could influence upfront treatment decisions for patients referred for radiation therapy. Potential changes include modified sequencing of systemic therapy and radiation, consideration of surgical options given higher complication risk in patients with ILD, or selecting systemic therapy alone in certain biomarker-defined cases until progression.
Title: Association of artificial intelligence-screened interstitial lung disease with radiation pneumonitis in locally advanced non-small cell lung cancer
Authors: Hannah Bacon, Nicholas McNeil, Tirth Patel, Mattea Welch, Xiang Y. Ye, Andrea Bezjak, Benjamin H. Lok, Srinivas Raman, Meredith Giuliani, B.C. John Cho, Alexander Sun, Patricia Lindsay, Geoffrey Liu, Sonja Kandel, Chris McIntosh, Tony Tadic, Andrew Hope
Read full article.
Leave a Reply