Data61 using AI and gamification to diagnose mental health patients

It will help accurately detect the difference between depression and bipolar.
Written by Aimee Chanthadavong, Contributor

Data61, the innovation arm of Commonwealth Scientific and Industrial Research Organisation (CSIRO), has used artificial intelligence (AI) and gamification to help psychiatrists and other clinicians accurately diagnose patients with mental health disorders and help improve overall mental health research.

Speaking to ZDNet during D61+ Live in Sydney on Wednesday, lead author of the research Amir Dezfouli said the idea for the research brings together his two areas of specialty: neuroscience and AI. 

"We know from neuroscience that most of the mental health disorders affect how we make decisions. One of the easiest ways to assess that is to complete a simple task or – in this case – a simple computer game, which allows us to record a patient's behaviour," he said.

"We then use machine learning AI to analyse this complex data set. This gives us an idea about the underlying pathology and gives us a diagnosis of mental health disorders."

He noted how currently, when it comes to diagnosing depression and bipolar, "there is a 60% chance of misdiagnosis of bipolar as depression and most go undiagnosed for more than 10 years".

"We hope with AI we can help clinicians with that and give them more information and help them to be more accurate," Dezfouli said.

"The other thing is some mental disorders go undetected for a long time. For example, OECD, it's something around 17 years. We are also hoping with AI, we can give people more personalised treatments."

Dezfouli explained the computer game that has been designed presents individuals with two choices and tracks the behaviour of how individuals respond to the choices.

"It shows the patterns of how we decide between [the choices]. Do you go between them, do you stick with one, and are able to figure out which is better or not," he said.

The study included 101 participants: 34 with depression, 33 with bipolar disorder, and a control group of 34 subjects.

Following the study, which has been selected to be presented at the 2019 Conference on Neural Information Processing Systems (NeurIPS) in Vancouver in December, Dezfouli is optimistic the gamification platform could be made commercially available.

"Eventually, we hope to provide it as a service or an online website that can help clinicians help solve the problem," he said.

Related Coverage

Editorial standards