I'm sorry, that's not going to happen. You could make something identify 10 or more symptoms, but no way will you get a five pound something diagnose any ten diseases with 100% accurtacy -- the diseases would have to be completely diagnosable from the surface of the patient, and even then, i bet you are going to need the software that i talked about in original post that can parse pictures to identify ailments. I don't get what you guys are saying: This contest is about producing a better medical swiss army knife -- if that's what's keeping those poor refugees from getting treated, then ok, go for it. but i think a lot more people could be helped by focusing on artificial diagnostic intelligence, using something like watson.