Voorspellen welke patiënt een psychotische stoornis krijgt of van te voren bepalen welke behandeling het effectiefst is? Het zijn zomaar wat grepen uit wat de toepassing van kunstmatige intelligentie in de psychiatrie kan opleveren. Toekomstmuziek? Of kunnen machinelearningmodellen de zorg nu al verbeteren?
The clinical application of neuroimaging for psychological complaints has so far been limited to the exclusion of somatic pathology. Radiological assessment of brain scans usually does not explain the psychological symptoms. However, that does not mean that psychological symptoms have no neurobiological basis. Hope has therefore been placed on functional MRI, which measures the activity of the brain. However, this has not yet resulted in clinical applications. A multivariate approach using machine learning analysis now appears to be changing this. Recent studies show that machine learning analysis of functional as well as structural MRI images can also provide diagnostic, prognostic and predictive biomarkers for psychiatry. Larger studies are needed to develop clinical applications, such as clinical decision support systems to support personalized treatment choices.
Conflict of interest and financial support: potential conflicts of interest have been reported for this article. ICMJE forms provided by the authors are available online along with the full text of this article.