The other day, many psychiatrists believed they would be able to make a precise diagnosis of schizophrenia at a glance after well trained. An experienced psychiatrist can sense the extraordinary atmosphere in a person with schizophrenia when facing him. Praecox-gefuehl was the most powerful diagnostic criterion of schizophrenia. Ridiculous? It is true that many great textbooks described Praecox-gefuehl even in the 20th century.
Of course, Praecox-gefuehl is a myth. It has been proven not to be trustworthy to identify schizophrenic patients.
On the other hand, some researchers are eager to identify some facial signs in psychotic patients. Dilated pupils, flat cheek, and less locomotive mouth were suspected to have a specific psychiatric symptom. There was a study to evaluate the mental status of a client with the tone of voice. However, these attempts had not shown a positive outcome.
And just recently, Google revealed a surprising result.
Facial recognition is a modern technology to distinguish one person from others. The preciseness of this technique is now so good that many people make use it for unlocking their smartphone. But it is still not perfect. In a recent incident, the program misidentified a black woman's face as a gorilla.
And now, researchers is utilizing deep learning to prove the existence Praecox-gefulhe. Millions of pictures of schizophrenic patients were recorded to be a subject to be learned by a supercomputer. If this trial is successful, a camera will be able to diagnose schizophrenia more precisely than human psychiatrists even without make a conversation with the patients.
Thus, I will lose my job. I hope this major trial will result in a failure.
No comments:
Post a Comment