By Aine Cryts

Radiologists are in a unique position to lead the development of artificial intelligence (AI) in medicine. That’s according to Ajay Kohli, MD, a radiology resident at the University of Pennsylvania Health System in Philadelphia, who credits the radiology specialty’s early embrace of technology in diagnosing and treating patients.

“From applying advancements in electromagnetism within modern-day MRs to image a wide range of diseases to using images for minimally invasive procedures, radiology has been at the forefront of modern medicine [in its use of technology],” he tells AXIS Imaging News.

Kohli points to specialties including gastroenterology, oncology, pathology, and surgery that are considering the use of AI. Radiologists can help lay the foundation for the future of AI, says Kohli, who notes that he doesn’t speak on behalf of the University of Pennsylvania Health System.

AXIS Imaging News recently discussed the role of radiologists in AI with Kohli. Here’s what he had to say.

AXIS Imaging News: What impact can you have as a practicing clinician and a trained radiologist?

Ajay Kohli: A radiologist typically goes through six years of advanced training—-all of which occur after four years of medical school and, commonly, three to four years of college. We’re trained in the complexities of a wide variety of pathologies and have to speak to several different subspecialty physicians at the level of their training. Thus, it’s important that radiologists are familiar with all the disease processes that subspecialty physicians typically see. We also need to understand how imaging findings will change treatment management.

Medical knowledge alone is a huge asset for anyone striving to bring AI to healthcare. In addition, most radiologists are experts in the clinical applications of the machines within which they work. With their experience interpreting various image types, such as x-rays, ultrasound, CT, and MR, radiologists can identify areas where AI algorithms are struggling to gain momentum. Take, for example, an algorithm to identify a fracture or an image artifact.

In addition, a growing number of radiologists are experts in technical innovations or healthcare management. These radiologists can help support the appropriate deployment of an AI product within the appropriate hospital setting. It’s no surprise that nearly all successful AI companies working in medical imaging are utilizing the experience of radiologists.

AXIS: What’s your advice for a radiologist who wants to get involved in AI or another technology project?

Kohli: Typically, radiologists help make important decisions within the hospital infrastructure, particularly when it comes to imaging equipment. Just as radiologists have led the way in determining which imaging equipment to purchase and adapt to their workflow, we’ll need to do the same for AI algorithms.

Vendor management, particularly of technology solutions, is one of the most important skills in healthcare; a physician who can achieve that will be invaluable to the healthcare system. All physicians, and particularly radiologists, should continue to play an active role in determining which AI vendors are chosen and implemented in their hospital system. Just as important? Radiologists must work closely with AI vendors to ensure that these products are useful in their workflow.

AXIS: What else should radiologists know about getting involved in AI or other technology projects?

Kohli: The financial cost, and the associated risk, of starting an AI company is quite high. Few, if any, radiologists are interested in starting their own companies or actively spending many hours each week advising existing companies. That said, Pittsburgh-based Bold Brain Ventures is reducing radiologists’ risk and time investment, while also ensuring that they have a role in shaping and benefiting from the coming AI revolution.

Bold Brain Ventures is an investment fund started by radiologists with investment experience. It’s empowering radiologists to help lead the AI revolution in radiology. The options for radiologists include investing as individuals or groups in a diversified portfolio of AI companies. This allows radiologists to take a financial stake in their area of expertise. The Bold Brain Ventures team also helps radiologists engage with AI developers and other stakeholders in this sector so that AI solutions can be developed and adopted in an ethical, appropriate, and commercially viable fashion.

[Kohli says that he has no financial or business relationship with Bold Brain Ventures.]

AXIS: Why is it important to discuss radiologists’ involvement in AI and technology?

Kohli: There will be many solutions in AI, specifically, and digital healthcare, in general. These platforms will draw the attention of healthcare leaders who may not have a background in medical imaging or, worse, may not value the expertise of the radiologist. It’s imperative that radiologists continue to take an active role in AI and digital health endeavors and lead the way in bringing the AI revolution to medicine.

Aine Cryts is a contributing writer for  AXIS Imaging News