Dr Borgen on the Use of AI to Support Treatment Decision-Making in Oncology

Patrick I. Borgen, MD, discusses how advances in artificial intelligence may influence

“AI is an inescapable reality for all of our futures. It’s not going to put us out of a job. It’s going to make our job different.”

Patrick I. Borgen, MD, chair, Department of Surgery, Maimonides Medical Center; head, Maimonides Breast Center, Maimonides Cancer Center, discusses how advances in artificial intelligence (AI) may influence the future of oncology practice.

AI is becoming an integral part of oncology care, Borgen begins. Currently, oncologists use AI for practice management tasks, such as booking and managing patient appointments, he says. However, its potential uses in treatment decision-making support are of interest in the oncology community, especially the breast cancer field, he notes. At the 2024 San Antonio Breast Cancer Symposium, discussions highlighted the complexities of managing HER2-positive breast cancer, where there are approximately 5 individualized treatment options at each of the 25 steps for patient care, he explains. Each decision is complicated by the emergence of clinical trials, underscoring the growing need for advanced decision-support systems that AI may be able to provide, according to Borgen.

The integration of AI tools into oncology is inevitable, but its implementation in the field must prioritize safety, HIPAA compliance, and human oversight, Borgen emphasizes. This technology is useful in controlled scenarios but not yet reliable for fully autonomous operation, he states. AI will not replace human expertise but will augment it by helping to manage the volume of clinical trials, treatment subtleties, and disease nuances across oncology, he reports.

The future success of AI in oncology lies in its careful application, Borgen says. AI tools will play a central role in matching treatments to specific diseases, allowing oncologists to navigate the complexities of personalized medicine more effectively, he explains. However, human interaction with AI will remain essential to validate its outputs and maintain control over its use.