Machine learning helps diagnose psychosis and depression
Technology has been developed that can more accurately distinguish between symptoms of psychosis and depression.
Created by researchers at the University of Birmingham, the technology could help improve diagnosis of both conditions, potentially leading to better treatments for patients.
Depression and psychosis are very rarely experienced as standalone conditions, with many people feeing symptoms of both at the same time.
The complex relationship between the two issues can cause problems with diagnosis, which many doctors try to solve by providing a 'primary' diagnosis with secondary symptoms.
However, this does not always provide an accurate reflection of an individual's experience, resulting in difficulty finding appropriate care and treatment.
For example, an individual diagnosed with psychosis as a primary condition may receive treatment geared toward treating hallucinations or delusions, with a lesser focus on treating depressive symptoms.
The University of Birmingham researchers have attempted to solve this problem using machine learning – a branch of artificial intelligence.
By analysing data from questionnaires, clinical interviews and MRI scans from 300 patients, the team were able to identify people with 'pure' diagnoses of either psychosis or depression, i.e. they had one condition with no symptoms of the other.
These 'pure' models were then applied to patients with symptoms of both conditions using machine learning to provide highly individualised psychological profiles.
When doing so, the team found that patients with depression as a primary condition were more likely to receive an accurate diagnosis.
At the same time, patients with psychosis as a primary condition had symptoms that were more associated with depression, suggesting that depression could play a larger role in psychosis than previously thought.
“The majority of patients have co-morbidities, so people with psychosis also have depressive symptoms and vice versa”, says lead study author Paris Alexandros Lalousis. “That presents a big challenge for clinicians in terms of diagnosing and then delivering treatments that are designed for patients without co-morbidity."
The challenges are not related to misdiagnosis, says Lalousis, but more to do with current diagnostic categories not accurately reflecting "clinical and neurobiological reality."
“There is a pressing need for better treatments for psychosis and depression, conditions which constitute a major mental health challenge worldwide. Our study highlights the need for clinicians to understand better the complex neurobiology of these conditions, and the role of ‘co-morbid’ symptoms; in particular considering carefully the role that depression is playing in the illness”.
To read the full study from the University of Birmingham, click here.