It would be interesting to see if there was a concurrence of erroneous diagnoses on a case by case basis and whether doctors tended to fail by erring on the false positive side as opposed to the computer or was it the other way round.