Interview with Regina Barzilay, distinguished professor for AI and Health at MIT (USA).
She is also an AI faculty lead for Jameel Clinic, an MIT center for Machine Learning in Health.
What is your view on the future of artificial intelligence in medical imaging for the next 10 years?
Regina Barzilay: Many of today’s successful applications of artificial intelligence (AI) in medical imaging focus on the automation of tasks that radiologists can do, for example cancer detection. As the field matures, I hypothesise that AI models will be able to answer many questions that are challenging for physicians such as: “What is the risk of getting a future disease?”, “What is the efficacy of a certain treatment?” and “How is a certain disease going to progress?”
Do you think that there should always be a human in the loop if we use AI in medical imaging?
Regina Barzilay: I cannot imagine healthcare without a human physician in the loop. While AI imaging tools can certainly bring many benefits, they are not perfect. Their best utilisation will depend on how well they are integrated into the clinical pipeline. It is up to human experts to design safe and effective protocols. I predict that as these imaging models continue to develop, a lot of low-level tasks will be delegated to AI, but the ultimate responsibility for clinical decisions will remain with physicians.
What obstacles are in the way of this view of the future?
Regina Barzilay: To bring AI-imaging tools into a clinical setting, it is essential to have the agreement of physicians. Today, AI is not part of the curriculum in most medical schools. This lack of background makes it challenging for physicians to adopt and trust this new technology.
Another important concern relates to the current privacy laws that govern data sharing. Models trained on a limited patient population are unlikely to generalise to a wider population. This can be particularly detrimental for minority groups, which may not be adequately represented in individual health centres.