How to Make Artificial Intelligence a Success in Medicine
Understand artificial intelligence (AI) and how its application can resolve your business challenges, advises AI executive Tushar Mehrotra. But don’t expect it to meet those challenges, if you don’t have the right talent onboard.
At the annual meeting of the Healthcare Information and Management Systems Society (HIMSS) this week in Orlando, Mehrotra will describe the kinds of talent that are critically important to the successful application of AI in health care and how organizations can get them. In a podcast published by Imaging Technology News, he explains how this talent might be acquired as well as the other steps needed to make AI a success in medical applications.
Mehrotra is senior vice president of analytics at Optum, which uses AI to improve patient experiences, as well as reduce health care costs. Other organizations can use AI in much the same way, Mehrotra explained in the podcast. Paramount to its successful use in healthcare, and particularly in medical imaging, is clearly understanding the issues that it can be leveraged to help resolve.
5 Steps to Successful AI
In the podcast, Mehrotra details five steps for the successful application of AI. The first is to develop a strategy for the use of AI. This entails working with the business and technology leadership of your institution to identify, and exactly characterize the issues, you want to resolve.
The second is to educate your organization’s business leaders about recent developments in AI. Separating fact from fiction — and convincing leaders that the fiction is just that — may not be as easy as it sounds. “You hear a lot of buzz … there are a lot of myths,” Mehrotra said in the podcast. But it is essential for leaders to understand clearly “what AI can and cannot do,” he said.
The third step comes after you have gotten the executive team onboard with the use of AI and have mapped out a strategy that describes how AI will be used. This step, Mehrotra said in the podcast, is the evaluation of AI technologies “so you have a sense of what to invest in.”
The fourth concerns getting the right talent for the job. This talent may be acquired by the hiring in-house staff. Or it might be acquired by partnering with other firms. Regardless of the route chosen, the talent must be integrated into your organization “so (it) is not working in a silo,” he said.
The fifth step is ensuring proper data curation. “This is going to be an incredibly important capability,” he said. Your organization must be able to aggregate, enrich and clean really, really large data sets from disparate data sources.” This last step relates directly to having the right talent onboard.
What Makes Medical Imaging a Prime Application Area
Medical imaging is “a pretty intense area for researchers and solution developers,” Mehrotra noted in the podcast. Some of the most substantial applications of AI in medical imaging have involved the interpretation of images. Exemplifying these, he said, are AI programs aimed at the interpretation of lung and liver lesions visualized by CT scans.
The reason is the extraordinary consumption of time that interpretation requires. Business use cases in medical imaging have a strong operational element, he noted. Consequently, a lot of attention is spent on “processes that take a lot of time.”
For those interested in using AI to resolve business use cases in health care, Mehrotra urges caution. The uninitiated must be careful not to “drown in the buzz,” he said in the podcast.
“Get an understanding of what artificial intelligence is,” he explained. “Understand what your specific business challenges are and how AI can be a vehicle to address those challenges.”
In the podcast, he advised prospective users of AI to be bold in their experimentation with AI. And to be sure to invest in talent:
“That will be a true differentiator for you and your business going forward.”
Greg Freiherr is a contributing editor for Imaging Technology News (ITN). Over the past three decades, Freiherr has served as business and technology editor for publications in medical imaging, as well as consulted for vendors, professional organizations, academia and financial institutions.