Purpose of review Disturbances in communication are a hallmark feature of severe mental illnesses. Recent technological advances have paved the way for objectifying communication using automated computerized semantic, linguistic and acoustic analyses. We review recent studies applying various computer-based assessments to the natural language produced by adult patients with severe mental illness. Recent findings Automated computerized methods afford tools with which it is possible to objectively evaluate patients in a reliable, valid and efficient manner that complements human ratings. Crucially, these measures correlate with important clinical measures. The clinical relevance of these novel metrics has been demonstrated by showing their relationship to functional outcome measures, their in-vivo link to classic ‘language’ regions in the brain, and, in the case of linguistic analysis, their relationship to candidate genes for severe mental illness. Summary Computer-based assessments of natural language afford a framework with which to measure communication disturbances in adults with severe mental illnesses. Emerging evidence suggests that they can be reliable and valid, and overcome many practical limitations of more traditional assessment methods. The advancement of these technologies offers unprecedented potential for measuring and understanding some of the most crippling symptoms of some of the most debilitating illnesses known to humankind.
Higher rates of non-right-handedness (i.e. left- and mixed-handedness) have been reported in schizophrenia and have been a centrepiece for theories of anomalous lateralization in this disorder. We investigated whether non-right-handedness is (i) more prevalent in patients as compared with unaffected siblings and healthy unrelated control participants; (ii) familial; (iii) associated with disproportionately poorer neurocognition; and (iv) associated with grey matter volume asymmetries. We examined 1445 participants (375 patients with schizophrenia, 502 unaffected siblings and 568 unrelated controls) using the Edinburgh Handedness Inventory, a battery of neuropsychological tasks and structural magnetic resonance imaging data. Patients displayed a leftward shift in Edinburgh Handedness Inventory laterality quotient scores as compared with both their unaffected siblings and unrelated controls, but this finding disappeared when sex was added to the model. Moreover, there was no evidence of increased familial risk for non-right-handedness. Non-right-handedness was not associated with disproportionate neurocognitive disadvantage or with grey matter volume asymmetries in the frontal pole, lateral occipital pole or temporal pole. Non-right-handedness was associated with a significant reduction in left asymmetry in the superior temporal gyrus in both patients and controls. Our data neither provide strong support for ‘atypical’ handedness as a schizophrenia risk-associated heritable phenotype nor that it is associated with poorer neurocognition or anomalous cerebral asymmetries.
This workshop summary on natural language processing (NLP) markers for psychosis and other psychiatric disorders presents some of the clinical and research issues that NLP markers might address and some of the activities needed to move in that direction. We propose that the optimal development of NLP markers would occur in the context of research efforts to map out the underlying mechanisms of psychosis and other disorders. In this workshop, we identified some of the challenges to be addressed in developing and implementing NLP markers-based Clinical Decision Support Systems (CDSSs) in psychiatric practice, especially with respect to psychosis. Of note, a CDSS is meant to enhance decision-making by clinicians by providing additional relevant information primarily through software (although CDSSs are not without risks). In psychiatry, a field that relies on subjective clinical ratings that condense rich temporal behavioral information, the inclusion of computational quantitative NLP markers can plausibly lead to operationalized decision models in place of idiosyncratic ones, although ethical issues must always be paramount.