Algorithm helps in early detection of psychosis
Artificial intelligence (AI) could make a significant contribution to the early detection of mental illnesses. A combination of artificial and human intelligence can improve the prevention of psychoses in young people, reports the Max Planck Institute of the results of a current study.
The research team around Professor Dr. Nikolaos Koutsouleris from the Max Planck Institute for Psychiatry combined machine learning models that analyze clinical and biological data with the assessments of treating doctors for the study. This led to a significant improvement in early detection compared to a medical prognosis alone. The study results were published in the specialist magazine “JAMA Psychiatry”.
The risk of poor progress is often underestimated
Although the specialists make very precise predictions about positive disease courses, the frequency of bad courses that lead to relapses is often underestimated, reports the Max Planck Institute. The researchers therefore asked themselves whether the transition to psychosis in patients with a clinically high risk or recent depression can be improved on the basis of machine learning.
Improve prediction of psychosis
In 334 people with a clinically high risk or a recent onset of depression and 334 people as a control group, the research team investigated whether the combination of specialist assessment with a computer-based evaluation of all clinical, neurocognitive, imaging and genetic information could improve the prediction of psychoses.
“The lack of prognostic sensitivity of the clinicians, measured by a false-negative rate of 38.5%, was reduced to 15.4% by the sequential prognosis model,” the researchers report on the results. Only the combination of AI and specialist assessment has optimized the prediction. “This enables us to improve the prevention of psychoses, especially in young patients at high risk or with newly emerging depression, and to intervene in a targeted manner in good time,” emphasizes Professor Koutsouleris.
Decision aid for practice
“The algorithm does not replace treatment by medical specialists, rather it offers decision-making aid and gives recommendations as to whether it makes sense to carry out further examinations on an individual basis,” the Max Planck Institute continues. On this basis, for example, it can be decided at an early stage which patients need therapeutic intervention and which do not.
Integrate into the clinical workflow
“The results of our study can help drive a two-way and interactive process of clinical validation and refinement of prognostic tools in real early detection services,” summarizes Professor Koutsouleris. When predicting psychoses, an individualized prognostic workflow that “sequentially integrates the risk assessments from algorithms and clinicians” can lead to significant improvements, the researchers write. However, the proposed workflow must be subjected to a comprehensive validation before clinical implementation. (fp)
Author and source information
This text complies with the requirements of specialist medical literature, medical guidelines and current studies and has been checked by medical professionals.
Dipl. Geogr. Fabian Peters
- Nikolaos Koutsouleris, Dominic B. Dwyer, Franziska Degenhardt, et al: Multimodal Machine Learning Workflows for Prediction of Psychosis in Patients With Clinical High-Risk Syndromes and Recent-Onset Depression; in: JAMA Psychiatry (veröffentlicht: 02.12.2020), jamanetwork.com
- Max Planck Society: An algorithm for the early detection of psychoses (published December 23, 2020), mpg.de
This article is for general guidance only and should not be used for self-diagnosis or self-treatment. He can not substitute a visit at the doctor.