In the study, AI scanned the MRIs of close relatives of schizophrenia patients.

Researchers at the University of Alberta developed an AI tool for predicting schizophrenia from brain scans of healthy close relatives (such as children or siblings) of persons suffering from the mental disease. Out of 57 persons, the AI system could identify the 14 individuals that scored the highest in a schizophrenia self-test. (News-Medical)

Prevalence of schizophrenia

While on average a Canadian faces only a less than 1% risk of developing schizophrenia during their lifetime, that percentage goes up to 19% if they happen to be a first-degree relative of a schizophrenia patient.

The disease afflicts nearly 300,000 Canadians. It causes confused thinking, lack of motivation, disorganized speech, and hallucinations.

It is incurable. However, programs and treatments are available to help manage symptoms in the community and at work.

The AI tool only supports decision-making, not a replacement for a diagnosis

Sunil Kalmady Vasu, Study Lead Author and Senior Machine Learning Specialist, Faculty of Medicine & Dentistry, University of Alberta, said the evidence-based AI tool analyzes the neural signature in the brain. It has the potential to be more accurate than diagnosis using the patient’s symptoms. The latter can be subjective, according to Kalmady Vasu.

However, the tool cannot be used as a substitute for clinical diagnosis by a psychiatrist.

“The goal is for the tool to help with early diagnosis, to study the disease process of schizophrenia and to help identify symptom clusters,” said Kalmady Vasu.

Next steps

The researchers will now apply the tool to persons unrelated to schizophrenia patients but exhibiting traits of the disease.

They also plan to track any persons red-flagged by the AI tool to determine whether they develop the disease later in their lives.

Related Story:  AI Algorithm Diagnoses Alzheimer’s Disease Better Than Experts