A recent study has revealed that artificial intelligence (AI) can significantly aid physicians in making more effective decisions in cancer treatment. Nevertheless, it also underscores the difficulties faced in the collaboration between medical professionals and AI. This research, conducted by Moffitt Cancer Center alongside the University of Michigan, examined the use of AI in radiotherapy specifically for non-small cell lung cancer and liver cancer (hepatocellular carcinoma).
A recent study led by researchers from Moffitt Cancer Center, in partnership with investigators from the University of Michigan, indicates that artificial intelligence (AI) can assist doctors in enhancing their cancer treatment decisions. Yet, the research also brings to light the challenges in the collaboration between physicians and AI. The findings, published in Nature Communications, concentrated on AI-augmented radiotherapy for non-small cell lung cancer and hepatocellular carcinoma (liver cancer).
Radiotherapy is a widely-used cancer treatment technique that employs high-energy radiation to eliminate or diminish tumors. This study evaluated a treatment method referred to as knowledge-based response-adaptive radiotherapy (KBR-ART). This innovative approach utilizes AI to enhance treatment results by proposing adjustments based on how effectively a patient responds to the therapy.
The research indicated that when physicians incorporated AI into their treatment planning, they made more uniform decisions, thus minimizing discrepancies in their choices. However, the technology didn’t always sway their opinions; in some instances, physicians opted to base their decisions on their clinical experience and the specific needs of their patients rather than on AI’s recommendations.
In the study, doctors were asked to evaluate treatment options for cancer patients initially without any technological aid and then with AI assistance. The AI system created by the researchers analyzes patient information, such as medical images and laboratory results, to suggest modifications in radiation doses. While some physicians found the AI’s recommendations beneficial, others preferred to rely on their own expertise.
“Although AI provides valuable insights from complex datasets, the human element is essential in cancer treatment,” stated Issam El Naqa, Ph.D., chair of the Machine Learning Department at Moffitt. “Each patient is distinct, and physicians must consider both AI proposals and their own clinical judgment when making decisions.”
The researchers emphasized that for AI to be effective, doctors must possess confidence in the technology. Their findings revealed that physicians were more inclined to adhere to AI recommendations when they trusted its guidance. “Our research demonstrates that AI can serve as a powerful asset for healthcare providers,” noted Dipesh Niraula, Ph.D., an applied research scientist in Moffitt’s Machine Learning Department. “However, it’s crucial to remember that AI functions best as an aid rather than a substitute for human expertise. Physicians contribute their knowledge and experience, while AI offers data-driven insights. Together, they can devise improved treatment plans, but this collaboration necessitates trust and effective communication.”
The authors of the study aspire for their discoveries to foster improved integration of AI tools and collaborative dynamics, aiding physicians in devising more tailored treatment plans for cancer patients. They also intend to further explore how AI can assist in other medical specialties.
This research received funding from the National Institutes of Health (R01-CA233487).
