Virtual PTSD Diagnostic Model

 Alberta: Eli is the name of a virtual computer model. This female model can talk to patients and tell them very accurately based on the text whether she is healthy or suffering from post-traumatic stress disorder (PTSD).

Scientists at the University of Alberta have enabled the model through machine learning to enable PDSD, a post-traumatic stress disorder, and stress disorder, with 80% accuracy from hidden gestures in the words people say. Can diagnose.

The reason is that there is a huge amount of data on the back of it so that the software can learn Harpal and move forward and can understand the psychological state of the person sitting in front.

It is hoped that such software will be used in the identification, treatment and potential psychological counseling of patients at home. Thus another door of telehealth is open in which computer models will be able to take the initial data and give their explanation.

For this project, Jeff Swalla, a doctoral student at the University of Alberta, has taken extensive data from the USC Institute for Creative Technologies and provided training on the software.

It included people's writings, statements and even tweets. Then they were ranked. Positive and negative phrases and words were observed in this process. There were also phrases and statements that people with depression and stress usually say over and over again.

The study looked at the text of a total of 250 interviews with 188 people, 87 of whom were diagnosed with PTSD by doctors.

A model named Eli was then prepared for the interview who skillfully talks to the person sitting in front of her or to the patient in such a way that the emotional aspects and feelings of her personality are best revealed. During this time she gives the patient a PTSD or fitness number.

But this is not an easy task. Sometimes the person sitting in the front does not open like a closed fist and he says vague phrases or words which makes it difficult to guess but it has been improved by adding more patient data.

In the next step, the sounds, movements and gestures of the machine model's accent will also be included in the study and test.