<html><body><div style="color:#000; background-color:#fff; font-family:times new roman, new york, times, serif;font-size:12pt"><div>I'm going off information I read in the BMJ in the 1990s, but back in the early days of computer-assisted diagnosis there were a few quick wins where computer diagnosis appeared to be outperforming humans. Analysis of the data showed this was occurring where humans would normally sit on the fence and be unsure if people had a certain symptom or sign, or if they were unsure if a critical threshold had been passed. The computer program was insisting the humans make a choice to move along the decision tree, so people were taking greater effort to look for signs or to take measurements carefully, and then actually make a firm decision.</div><div><br><span></span></div><div style="color: rgb(0, 0, 0); font-size: 16px; font-family: times new roman,new york,times,serif; background-color: transparent; font-style:
normal;"><span>Tom</span></div></div></body></html>