What’s the source for 10% specificity? I couldn’t find any article that said this was the case. In fact the only article that mentions any numbers is written by Quartz and says this: “Cardiologists were able to read 90% of the total readings, although about 10% of them were unreadable.” That’s not false positive.