Algorithmic Classification of Psychiatric Disorder–Related Spontaneous Communication Using Large Language Model Embeddings: Algorithm Development and Validation

This study introduces an innovative use of LLMs in psychiatry, showcasing their potential to objectively examine language use for distinguishing between different psychiatric disorders. The findings highlight the capability of LLMs to offer valuable insights into the linguistic patterns unique to various conditions, paving the way for more efficient, patient-focused diagnostic and monitoring strategies. Future research should aim to validate these results with clinically confirmed populations and investigate the implications of comorbidity and spectrum disorders.