Sunday, August 30, 2015

IBM e investigadores suramericanos desarrollan software para … – Hipertextual

Image: agsandrew - Shutterstock

: agsandrew – Shutterstock

Schizophrenia is a chronic mental disorder that can develop at any time of life and, though it has various manifestations, generally characterized by the presence of hallucinations, emotional distance, features, and paranoid delusions. Traditionally the early detection of this disorder requires a professional eye supported by interviews and application of psychological tests. However, diagnostic process is complicated , subjective and fallible; two experts could observe the same patient and reach different conclusions. In fact, unless the patient to attend regular appointments with the psychiatrist, the chances are that their condition passes unnoticed until their first psychotic break is triggered.

The people susceptible to schizophrenia tend to have slight differences in the structure of his speech
However, thanks to research conducted jointly by Columbia University, the University of Buenos Aires and IBM, this could change. One of the main characteristics of schizophrenia is the loss of coherent thought, manifested in expressions disorganized speech, or put simply, the patient begins to say crazy things. The interesting thing is that even without the disorder in the present, susceptible to the disease tend to have slight differences in the structure of his speech. The researchers relied on this fact to develop software to detect schizophrenia , which uses the speech pattern recognition to detect these small variations and make a diagnosis with 100% efficiency.

Already in the past had used this type of software for compare patterns of speech the relatives of schizophrenics and healthy, sustained the notion that this disease has a strong genetic component, but this new work aims to develop a software that can be applied in the clinical setting, giving the specialists an objective method to make the diagnosis. When we consider that traditional methods are 70% effective (7 out of 10 patients are correctly diagnosed), these seem to be great news, especially when you consider that the software has the ability to learn using preliminary tests to refine your criteria detection.

 Image: goa Novi - Shutterstock

Image: goa Novi – Shutterstock

However, there is still a long way to go, as the research sample was only 34 subjects (23 men and 11 women), enough to detect a pattern, but it would be necessary to repeat the work with a much larger sample and different socio-cultural contexts, to validate the procedure. Additionally, the software has only been designed to work in English, and when we consider the number of languages ​​and regional variations thereof, can realize the hard work that psychiatry ahead.

Of course, it is also important to remember the ethical considerations of this finding, although we know that there is a connection between genetics and schizophrenia, still do not know why some vulnerable people never develop the disorder, catalog or someone schizophrenic potential schizophrenic could help in many cases, but it could also be labeled so many others and allocate them to a life of stigma and unnecessary prescriptions. This type of tool have the same moral dilemma Minority Report, does the fact that a person may commit a crime in the future automatically make him a criminal?.

Apart from this, work is deeply encouraging, not only to patients who may suffer from this disorder but for psychiatry in general; a frequent criticism of this sub-discipline is the old and questionable validity of some of its instruments, jobs like these researchers are a jumper between the computer and the nebulous realm of the psyche , something 20 years ago it would have been relegated to the pages of a science fiction novel. These researchers hope that further work is beyond schizophrenia and that new technologies can be used for other disorders such as depression, obsessive compulsive disorder or autism.

LikeTweet

No comments:

Post a Comment