Is Artificial Intelligence Hitting a Healthcare Wall?
Robin Young • Wed, July 5th, 2017
Evan Sweeney, senior editor at Fierce Healthcare, noted in a recent article that tech giants like IBM, General Electric and Google may be hitting the wall when trying to apply artificial intelligence (AI) tools to the problems of diagnosing and treating patients.
“Technology giants like IBM, General Electric and Google have been eager to capitalize on AI advancements that could improve medical care. But for AI to make its mark, it will have to overcome a fundamental flaw that has clung to healthcare for decades: access to patient data.”
The problem, which every, single physician in America confronts every, single day…is figuring out right answers from wrong answers. Surprise…biology and psychology…is complicated.
Said Sweeney in his June 27 blog, “Access to data is just one piece of the puzzle. Machine learning tools also need to be fed data that differentiates right answers from wrong answers. For particularly complex conditions, that kind of easily digestible data might not exist.”
In short, there may not be a readily identifiable right or wrong answer. In fact, in medicine, the solution to either a diagnostic problem or a treatment problem may be a multi-year, incremental process.
"In a specialized domain in medicine, you might need experts trained for decades to properly label the information you feed to the computer," Thomas Fuchs, a computational pathologist at Memorial Sloan-Kettering, told MIT Technology Review.
Continued Sweeney: “Researchers have raised this issue before, with some arguing the next generation of machine learning software needs to capture a “richer clinical picture” by tapping into physician-generated crowdsourced data. Data scientists have also traced AI’s practical shortcomings to an inability to access robust data, like social determinants of health, from EHRs [electronic health records].”
Office of the National Coordinator for Health IT's Deputy National Coordinator Jon White, M.D., tweeted from an event hosted by SMART Health IT, pointing to the potential complications of using bad data for machine learning, and hinting that improving datasets may be a role the government can take on.
As Manish Kohli, a healthcare informatics expert with the Cleveland Clinic, told MIT Technology Review, “healthcare has been an embarrassingly late adopter of technology.”
DePuy Synthes has touted its partnership with IBM’s AI platform Watson. Google is trying to apply AI to create a smart OR environment with its company VERB.
In Sweeney’s view, the low hanging fruit for AI in healthcare is imaging. “AI could have the most success in image-based specialties where the data is more transparent and the diagnoses are less ambiguous.”