Canadian journal of psychiatry. Revue canadienne de psychiatrie
-
Neuroimaging research has substantiated the functional and structural abnormalities underlying psychiatric disorders but has, thus far, failed to have a significant impact on clinical practice. Recently, neuroimaging-based diagnoses and clinical predictions derived from machine learning analysis have shown significant potential for clinical translation. This review introduces the key concepts of this approach, including how the multivariate integration of patterns of brain abnormalities is a crucial component. ⋯ We propose that longitudinal outcomes, such as early diagnosis and prediction of treatment response, offer definite opportunities for progress. We propose that efforts should be directed toward clinically challenging predictions in which neuroimaging may have added value, compared with the existing standard assessment. We conclude that diagnostic and prognostic biomarkers will be developed through the joint application of expert psychiatric knowledge in addition to advanced methods of analysis.
-
It is unknown whether evidence-based, in-hospital processes of care may influence the risk of criminal behaviour among patients with schizophrenia. Our study aimed to examine the association between guideline recommended in-hospital psychiatric care and criminal behaviour among patients with schizophrenia. ⋯ High-quality, in-hospital psychiatric care was associated with a lower risk of criminal behaviour after discharge among patients with schizophrenia.