To correctly identify patients with a combination of psychotic and depressive symptoms, researchers from the University of Birmingham recently developed a method to use machine learning to do so.
The results of the research were published in the journal ‘Schizophrenia Bulletin’.
Patients with depression or depression rarely experience symptoms of one or the other illness. Historically, this has led to mental health clinicians giving a diagnosis of ‘primary’ illness, but with secondary symptoms. Making a correct diagnosis is a major challenge for clinicians and studies often do not reveal the complexity of individual experience or indeed neurobiology.
For example, clinicians experiencing psychosis, for example, would often view depression as a secondary illness, with an impact on treatment decisions that focus more on symptoms. mindfulness (e.g. hallucinations or delusions).
A team at the University of Birmingham Institute of Mental Health and the Center for Human Brain Health, worked with researchers from the PRONIA consortium to study the feasibility of using machine learning to create highly accurate models of ‘pure’ forms of the two diseases and used these to examine the diagnostic accuracy of a group of patients with mixed symptoms. Their findings are published in the Journal of Schizophrenia.
“Most patients have co-infections, so people with depression and vice versa have depressive symptoms,” explained lead author Paris Alexandros Lalousis.
Read also: McDonald’s is considering selling part of Dynamic Yield Startup
Lalousis said, “That is a major challenge for clinicians in terms of finding and then delivering treatments designed for patients without comorbidity. It is not that patients are misdiagnosed, but our current diagnostic areas do not reflect clinical and neurobiologic facts ”.
The researchers analyzed questionnaire responses, in-depth clinical interviews, and data from structural magnetic resonance imaging from a group of 300 patients participating in the PRONIA study, an EU-funded cohort study taking place across seven European research centers.
Within this group, the researchers identified small subgroups of patients who could be classified as either suffering from dementia with no symptoms of depression, or from depression with no symptoms of depression.
Using this data, the team identified machine learning models of ‘real’ depression and ‘pure’ mindfulness. They were then able to use machine learning techniques to apply these models in patients with symptoms of both diseases. The goal was to build a true disease profile for each patient and test against their diagnosis to see how accurate it was.
The team found that while patients with depression as a primary illness were more likely to be diagnosed correctly, depressed patients with depression usually had symptoms that usually looked low depression. This may indicate that depression plays a larger role in the illness than previously thought.
Lalousis said, “Better treatments are desperately needed for depression and depression, conditions that are a major mental health challenge worldwide. Our study highlights the need for clinicians to better understand better understanding of the complex neurobiology of these conditions, and the role of ‘co-morbid’ symptoms; in particular, carefully considering the role of depression in the illness “.
Lalousis noted, “In this study, we have shown how the use of sophisticated machine learning algorithms that take into account clinical, neurocognitive, and neurobiologic features can aid our understanding of complexity. mental illness. “
Lalousis also said “in the future, we believe that machine learning could become an essential tool for proper diagnosis. We have a real opportunity to develop data-based diagnostic methods – this is an area in which mental health Maintaining physical health and it is vital that we maintain that trend. “