Not necessarily. If you were running a casino and someone was able to significantly beat the odds on 34 of your 36 tables, would you assume they cheated at all

Not necessarily. If you were running a casino and someone was able to significantly beat the odds on 34 of your 36 tables, would you assume they cheated at all other tables but not those two, or just that other players at those tables were really good? It could be that 2 of those systems are actually better. It could also be that the state of the art in deep learning just isn't there yet for radiology. But this is a strong suggestion that research is being dramatically overstated, so I would get real careful with my investments in this area overall.