SEP 05, 2018 8:23 AM PDT

Can AI Algorithms Accurately Assess Mental Illness?

Mental illness is complicated because while the causes are rooted in brain chemistry, neurological injury or other mechanisms in the brain, what is most affected are the daily lives of patients and their families.

Living with mental disorders like major depressive disorder or bipolar disorder is difficult, especially since many patients do not find relief in medications. Talk therapy and cognitive behavioral therapy (CBT) are other ways to treat mental illness; however it’s very much a trial and error process that can be time-consuming and frustrating.

Researchers from Lawson Health Research Institute, The Mind Research Network, and Brainnetome Center, collaborated recently on a study that shows how artificial intelligence (AI) and a precisely created algorithm that can analyze brain scans. The technology is aimed at better classifying diagnoses in patients and helping to predict what the patient’s response to specific medications might be, based on their particular illness characteristics.

The study cohort included 78 adult patients from mental health programs at London Health Sciences Centre (LHSC) and who were mostly from the First Episode Mood and Anxiety Program (FEMAP.) The study began with 66 patients who had completed treatment for a definite diagnosis of either major depressive disorder (MDD) or bipolar type I. Type I bipolar disorder is a form of the disease that includes full manic episodes as well as depressive episodes. There were also an additional 33 healthy patients, with no history of mental illness, as a control group. All of the participants underwent functional MRI (fMRI) scans as part of the research.

The research team analyzed the scans from each group and found, not surprisingly, that there were distinct differences in brain networks including the default mode network which regulates self-reflection, and the thalamus which is a connector hub that links multiple cortical regions and is involved in arousal and alertness. This data was the basis for an AI algorithm that used machine learning to pinpoint whether a patient had MDD or bipolar I When the AI program was tested to see if it could accurately diagnose the correct disorder based on the scans, it scored an impressive 92.4% for accuracy.

The research team then took it up a notch and had the AI algorithm look at brain scans from a dozen other patients who had did not have a precise diagnosis due to overlapping symptoms and other factors. Patients with MDD will often respond well to antidepressants, while patients who have bipolar I will do better on mood stabilizing drugs. The AI algorithm correctly predicted which disorders were present and which class of medications would have the best patient response in 11 of the 12 difficult cases.

Dr. Elizabeth Osuch, a clinician-scientist at Lawson, medical director at FEMAP and co-lead investigator on the study, explained, Antidepressants are the gold standard pharmaceutical therapy for MDD while mood stabilizers are the gold standard for bipolar I. But it becomes difficult to predict which medication will work in patients with complex mood disorders when a diagnosis is not definite. Will they respond better to an antidepressant or to a mood stabilizer? This study takes a major step towards finding a biomarker of medication response in emerging adults with complex mood disorders,” says Dr. Osuch. “It also suggests that we may one day have an objective measure of psychiatric illness through brain imaging that would make diagnosis faster, more effective and more consistent across health care providers.”

The video below features Dr. Osuch as well as more information on the research, take a look.

Sources: R&D Mag, Lawson Research, Acta Pyschiatrica Scandanavica

About the Author
  • I'm a writer living in the Boston area. My interests include cancer research, cardiology and neuroscience. I want to be part of using the Internet and social media to educate professionals and patients in a collaborative environment.
You May Also Like
OCT 05, 2020
Earth & The Environment
The rising concern of aerosol particles
OCT 05, 2020
The rising concern of aerosol particles
A study from Colorado State University scientists provides insight into the resiliency of aerosol particles, particles f ...
OCT 14, 2020
Health & Medicine
Sanobiotec Produces Reliable THCV
OCT 14, 2020
Sanobiotec Produces Reliable THCV
Lithuania-based bio Sanobiotec has joined a band of biotech companies moving into the synthetic cannabinoid market and a ...
OCT 21, 2020
Drug Discovery & Development
New ALS Treatment Extends Life for Several Months
OCT 21, 2020
New ALS Treatment Extends Life for Several Months
Currently, there are only two approved medications to treat Lou Gehrig's disease (also known as ALS), a condition po ...
OCT 29, 2020
Cardiology
Treating Cardiovascular Calcification at the Source
OCT 29, 2020
Treating Cardiovascular Calcification at the Source
The term cardiovascular disease covers a broad array of health problems. Everyone tends to think of heart attacks or hyp ...
OCT 24, 2020
Genetics & Genomics
Cord Blood Samples Reveal More About the Genetics of Autism
OCT 24, 2020
Cord Blood Samples Reveal More About the Genetics of Autism
The activity of genes in our genome is controlled by many factors, one of which are chemical tags or structural changes ...
OCT 25, 2020
Health & Medicine
Over 4 in 10 People With MS Have Used Cannabis in the Last Year
OCT 25, 2020
Over 4 in 10 People With MS Have Used Cannabis in the Last Year
University of Michigan researchers have found that Americans with Multiple sclerosis (MS) are twice as likely as the gen ...
Loading Comments...