hmmm... unless I see a formal study/data thru some kind of scientific study, I am not convinced of the validity of this story. The thing with facial recognition is, we humans have a tendency to make errors in recognizing faces. If this 'psychiatrist' showed the picture to his/her clients, you're already priming them to try to recall something that they (think) they saw in a dream. Memory could be such a faulty thing. You prime a person using an ambiguous face like that and you will get confirmation of 'sightings'.
So for now, I call BS.
So for now, I call BS.