As I recently wrote here, Artificial Intelligence (AI) is creeping into healthcare, and particularlyradiologic services. What I didn’t realize then, but do now, is to what extent AI is already involved in your radiology services.
The Providence healthcare system has partnered with a company called Nuance for over a year. Recently, Nuance announced it was launching an “AI-Powered Precision Imaging Network” to assist radiologists and radiology departments in providing radiologic interpretations to patients in Washington and elsewhere.
From Nuance’s website, this AI “Network” works by having the images immediately uploaded for AI interpretation. Once interpretation is complete, depending on the results the particular study can be marked with urgency (such as when AI interprets a patient has a life-threatening condition) and moved to the front of the actual radiologist’s list of images to interpret. It also generates text of its interpretation, so that a radiologist may insert it directly into the patient’s imaging study report.
Nuance is marketing this new “Network” as beneficial to not only patients, but “financial performance” and “efficiency across the entire patient journey from screening through follow-up.” Given the cost of paying human beings to interpret imaging studies, presumably this new “Network” will allow radiologists to interpret more films in less time. The question is begged: when will AI take over such tasks completely? It seems to be a matter of when, not if.
From a medical malpractice perspective, it is often useful to know what certain people knew and when. Specific to a case hinging on a radiologic study and interpretation, if a radiologist (or radiology group) uses Nuance, and Nuance’s AI identified a serious issue needing prompt attention, that would be useful information if the reality was a significant delay in human interpretation of the images at issue.
What happens when the AI program “misses” a serious, life-threatening condition, a radiology group relies on the program to assess the acuity of each patient, and because of the delay the patient is harmed?
You can read more about this company, and this new “Network”, here.
