Question

@meisen C points out a potential situation where the cardiologist had higher scores than the program. We know the program correctly diagnosed more actual heart attacks, but we don't know if it perhaps over-diagnosed them. It may well note something as a heart attack that isn't. C gives us the possibility that the cardiologist is better equipped to notice when an issue isn't a heart attack, meaning that the cardiologist doesn't over-diagnose to the extent the program does. This possibility would weaken the argument that we should leave diagnoses to the program.