AI bests radiologists at predicting lung cancer treatment responses
New data suggest artificial intelligence-enabled tools often are more accurate than radiologists at predicting treatment responses.
Published in Frontiers In Oncology, the research suggests that AI’s knack for interpreting post-treatment imaging exams could help providers to make more informed decisions on how to best manage patients’ treatment. Doing so could lead to improved outcomes, especially for those battling lung cancer.
“Despite advances in targeted therapy and immunotherapy, many patients, particularly those with advanced-stage disease, continue to experience poor outcomes, underscoring the importance of early, accurate treatment response assessment to guide timely therapeutic decisions and avoid ineffective toxicity,” Nehemias Guevara Rodriguezm, MD, with the Division of Hematology, Oncology and Bone Marrow Transplant at Saint Louis University School of Medicine, St. Louis, and colleagues explained. “Artificial intelligence systems spanning radiomics pipelines and deep learning architectures can quantify high-dimensional image features and temporal changes beyond human perception, promising earlier and potentially more objective prediction of treatment response.”
To get a better idea of how AI’s predictive performance compares to that of radiologists in different contexts, the group conducted a meta-analysis of 11 studies published in PubMed, Embase, Scopus, Web of Science, and the Cochrane Library from their inception until April of 2025. The studies’ data, which spanned over 6,600 patients, were retrospectively evaluated by two reviewers who compared the performances of AI and readers based on their sensitivity, specificity, accuracy and risk differences.
Based on those metrics, AI outperformed radiologists, achieving a sensitivity, specificity and accuracy of 0.90, 0.80 and 0.90, respectively. Compared to radiologists, the AI tools achieved pooled sensitivity and specificity improvements of approximately 6% and 4%.
The difference in performance was most notable in PET and CT imaging. This finding is likely owed to the modalities’ alignment with AI pattern recognition capabilities, the group suggested.
“These findings represent a meaningful advancement in precision oncology, where even marginal gains in predictive accuracy can translate to substantial clinical benefits given the high stakes of treatment selection in lung cancer management,” the authors noted. “This paradigm shift from reactive to proactive treatment management represents a fundamental evolution in oncological practice, potentially improving both survival outcomes and quality of life for patients with lung cancer.”
The group acknowledged that more prospective data are needed before AI can reliably assist in treatment decisions. However, they believe the ability of AI to better predict how patients will respond to treatment will give providers an opportunity to be proactive, rather than reactive, in managing treatment decisions in the future.
Read more about the findings here.
In addition to her background in journalism, Hannah also has patient-facing experience in clinical settings, having spent more than 12 years working as a registered rad tech. She began covering the medical imaging industry for Innovate Healthcare in 2021.