Magnetic resonance imaging to predict the risk of relapse after the first psychotic episode?

A recent study has delved into the question of whether an MRI scan could predict a patient’s risk of relapse after a first psychotic break.

Combining the data obtained from the magnetic resonance imaging with the patient’s clinical data could make it possible to make fairly reliable estimates of the risk of relapse after a first psychotic episode. This is shown by the results of the study, led by researchers from the August Pi i Sunyer Biomedical Research Institute (IDIBAPS) of the Hospital Clínic in Barcelona, ​​together with the FIDMAG Foundation and the Mental Health Network Biomedical Research Center (CIBERSAM), in Spain.

The study was coordinated by Joaquim Raduà, head of the Imaging of Mood and Anxiety Related Disorders (IMARD) research group at IDIBAPS and CIBERSAM researcher, and Edith Pomarol-Clotet, director of the Hermanas Research Foundation Hospital (FIDMAG) and CIBERSAM researcher. The first author of the paper is Aleix Solanes, a researcher in the IDIBAPS IMARD group.

The discovery of the relationship between magnetic resonance imaging and mental disorders prompted imaging biomarker research in this area, but no relevant findings had been achieved to date. New artificial intelligence methods have reopened the possibility of creating tools that can help in clinical decision making.

“The fact of estimating this risk is important because this way we could personalize the treatment and increase the frequency of visits in high-risk patients, or avoid unnecessary treatments if the risk is low,” says Joaquim Raduà. “Until now, you could only tell if a patient was at high risk of relapse after multiple flares and relapses,” he adds.

The study involved 227 patients with a first psychotic episode from seven hospitals in Spain who were followed up for 2 years. The researchers designed a tool to assess whether MRI was capable of detecting patients at high risk of relapse after the first flare.

The results of the study demonstrate that the software was able to detect patients with a high risk of relapse. In fact, in empirical validation, the observed risk of relapse in patients detected as high risk was 4.5 times higher than in low risk patients.

Artificial intelligence analysis of MRI data could predict the risk of relapse after a person’s first psychotic break. (Image: Hospital Clínic / CIBERSAM)

This neuroimaging data-based tool to assess the risk of relapse after a first psychotic break is valuable and feasible. “Valuable because it allows adjusting the follow-up and treatment of each patient based on the determined risk. Feasible, since people with a first psychotic outbreak already have an MRI to rule out any organic disease in the brain; no additional examinations are required”, conclude the researchers.

The software developed for this study has been made freely available to the scientific community to allow groups to develop their own detection model. On the other hand, there is also a website to estimate the risk of relapse quickly and help replicate the model. But before being able to use this tool in clinical practice, the researchers stress that other studies need to replicate the results.

The study is titled “Combining MRI and clinical data to detect high relapse risk after the first episode of psychosis”. And it has been published in the academic journal Schizophrenia, from the Nature Group. (Source: Hospital Clínic / CIBERSAM)