New AI platform could help cancer survivors get personalized support between doctor visits

Researchers at Sylvester Comprehensive Cancer Center, affiliated with the University of Miami Miller School of Medicine, have devised a framework to develop AI tools for cancer survivorship and supportive care, as demonstrated in a proof-of-concept study published in Translational Behavioral Medicine. The Precision AI for Survivorship and Supportive Care framework aims to tailor evidence-based support for survivors coping with persistent side effects and challenges during the period between doctor visits.

According to Frank Penedo, the study's lead and director of Sylvester's Survivorship and Supportive Care Institute, AI could help extend supportive care beyond cancer center walls, fostering greater connection, information, and support for survivors while preserving in-person care and access to their healthcare providers. The framework leverages years of survivorship research and clinical infrastructure developed at Sylvester, utilizing data from over 37,000 patients through the My Wellness Check program, an electronic health record-integrated screening and triage system.

By evaluating data from a subset of 25,592 patients followed for 36 months, researchers developed machine-learning models that identify patterns linked to symptom severity and unplanned healthcare use, boosting predictive precision for adverse outcomes by more than 25%. This data and infrastructure underpin the development of a personalized supportive care strategy through an AI-enabled platform known as My Wellness Support.

The platform integrates patient-reported outcomes, clinical records, and behavioral information to pinpoint risk patterns and unmet needs, enabling the delivery of customized educational resources, symptom-management techniques, resources for practical needs, and tailored supportive care recommendations. The AI companion featured in the platform works alongside researchers and clinicians, monitoring trends, risk scores, and alerts through a dedicated dashboard.

The goal is to supplement, rather than replace, clinical interactions, offering patients access to trusted, evidence-based information when needed. The platform is designed to adhere to established clinical guidelines and safety protocols while ensuring human oversight. Researchers stress the importance of rigorous scientific validation, transparency, and continuous evaluation to ensure the safe and effective integration of AI into survivorship and supportive care.

The study's findings, led by Frank J. Penedo et al., highlight the potential of AI to enhance support for cancer survivors between appointments and may serve as a model for the responsible integration of AI technology in this domain.