While Elon Musk believes artificial intelligence (AI) is the biggest risk we face as a civilization, the technology is already proving its worth in the medical field.
The University of Alberta (UoA) has been collaborating with IBM to use AI and machine learning algorithms to quickly diagnose schizophrenia with 74 per cent accuracy. The technology is also being used to predict the severity of specific symptoms in patients – something that was not possible before, IBM says.
“Using AI and machine learning, ‘computational psychiatry’ can be used to help clinicians more quickly assess – and therefore treat – patients with schizophrenia,” says Guillermo A. Cecchi, principal research staff member in computational neuroscience at IBM Research, in a July 20 press release. “Computational psychiatry provides physicians with tools that enable them to objectively assess patients where most approaches had been subjective up until that point….For the first time, clinicians could be able to quantitatively determine the severity of common symptoms and even identify and measure the progression of the disease, as well as the effectiveness of treatment.”
IBM’s Alberta Centre for Advanced Studies (CAS) and UoA have been partners for more than a decade, and the two organizations conducted this schizophrenia research earlier this year using 95 test subjects – 46 patients with schizophrenia, and 49 patients without.
The goal of the study was to extend IBM’s research with the UoA, and connect the tech giant’s computer scientists with the university’s psychiatry and computer science department, giving them access to a much larger group of patients and data.
Approximately one in five adults in North America suffer from a mental health condition at some point in their lives, ranging from depression to bipolar disease to schizophrenia, but half of those with severe psychiatric disorders receive no treatment. And in the case of schizophrenia, there is no medical testing that can provide an absolute diagnosis.
Going forward, the team hopes to apply this to other diseases, such as Huntington’s, and extend this research model across larger groups of patients, according to Betakit.
