Real-Time AI Trial Demonstrates Potential to Accelerate Cancer Diagnosis and Treatment

A pioneering study conducted by researchers from the Icahn School of Medicine at Mount Sinai, Memorial Sloan Kettering Cancer Center, and collaborators has showcased how artificial intelligence (AI) can transform cancer care. The research involved a real-time, behind-the-scenes trial where AI analyzed live pathology samples, offering promising results in predicting genetic mutations crucial for personalized treatment.

Published in Nature Medicine, the study revealed that AI models could accurately forecast genetic mutations, such as EGFR in lung adenocarcinoma, directly from routine histopathology slides. These slides, stained with standard hematoxylin and eosin (H&E), are part of normal diagnostic procedures, making the approach highly accessible. By identifying mutations automatically, the AI could reduce reliance on expensive and time-consuming genetic tests, potentially decreasing the need for rapid genetic assays in some cases by over 40%.

The innovative AI system was carefully trained on extensive datasets derived from multiple institutions across the US and Europe, linking pathology slides with next-generation sequencing results. This setup allowed the AI to learn how to detect critical biomarkers from routine tissue images and to be integrated into standard clinical workflows.

"Our goal was to see if AI could extract vital genetic information from standard diagnostic slides, supporting faster and more efficient treatment decisions," explained Dr. Gabriele Campanella, the study's lead author. Although confirmatory genetic testing remains necessary, the AI's ability to flag likely mutation cases early on could optimize resource use and hasten the initiation of targeted therapies.

During the trial at Memorial Sloan Kettering, the AI analyzed patient samples in real time without clinicians viewing its predictions. The results confirmed the model's reliability in detecting EGFR mutations, suggesting that such systems could significantly streamline the diagnostic process.

The findings emphasize that AI can play a critical role in enhancing cancer diagnostics, making treatment planning swifter and more personalized. Researchers are now focused on expanding the system to identify additional biomarkers and to evaluate its performance in resource-limited settings where access to advanced genetic testing is limited. This approach holds promise for broader clinical adoption, aiming to improve patient outcomes worldwide.