Artificial intelligence (AI) and machine learning (ML)-based tools are rapidly becoming more mainstream. What does that mean for you, as an oncologist, and for your patients? Are you comfortable using AI decision models in your practice? If you’re hesitating to use these models, what concerns you the most?

NCI-funded researchers recently conducted a survey to help shed light on how oncologists feel about using AI. The survey captured information on how well oncologists understood AI and how confident they were at explaining it.

The results support AI’s place in clinical practice. Of the survey respondents, most reported that AI could help in diagnosing (95.6%), treating (89.2%), and managing side effects (60.4%).

However, respondents also expressed ethical concerns:

• 81.4% of the survey respondents felt patients should give consent for using AI-based models in treatment decisions.
• 90.7% felt AI developers should bear responsibility for medical-legal problems associated with AI use, although some reported the physician or hospital also should share in that responsibility.
• 76.5% reported feeling responsible for protecting patients from biased AI. But only 27.9% felt confident that they could spot bias.

What does this mean for the future of AI use in the clinic?

Corresponding author, Dr. Gregory Abel, of the Dana-Farber Cancer Institute, notes, “A large majority of the respondents felt they’d benefit from training on using AI for cancer care, but many didn’t know where to find those resources.”

“This underscores how important it is to give oncologists the information they need to assess AI’s effects on care decisions. By better understanding AI and how it works, they will also be in a better position to understand their responsibilities when using AI for clinical care,” he added.