In The Yuan
From my article in The Yuan, posted August 1, 2022
Medical errors cost the lives of some 250,000 people a year in the U.S. alone, calculates the BMJ, making them the third leading cause of death, with a price tag of some $20 billion a year. Research suggests AI can help cut this staggeringly high error rate. For example, AI systems have been shown in different studies to exceed physicians’ accuracy in diagnosing cancer in both pathology slides and imaging, and in detecting brain hemorrhages that lead to stroke, reducing both false positives and false negatives.
But even if AI systems get it wrong at a lower rate than physicians, they may still be seen as causing more problems than they solve. That’s because there is a well-established infrastructure designed to catch, analyze, prevent and correct conventional medical errors. But there is as of yet little such infrastructure for AI medical errors….Read more
David H. Freedman 
AI is used as an assist to physicians in diagnostic imagining, not to replace them. There is not an imaging company that I know of that is comfortable with their technology to have AI alone diagnose disease tissue.