<b>A transdiagnostic prefrontal fNIRS dataset from 1,228 participants during a verbal fluency task</b>

Mental disorders represent a profound global public health challenge, yet their diagnosis remains heavily reliant on subjective clinical assessments due to a lack of objective neurobiological biomarkers. Developing diagnostic technologies based on functional near-infrared spectroscopy (fNIRS) offers a promising avenue for facilitating objective, automated psychiatric screening. To address the critical scarcity of large-scale clinical data, a transdiagnostic fNIRS dataset was constructed from 1,228 participants during verbal fluency task. The cohort included 53 healthy controls and 1,175 patients diagnosed across four major psychiatric categories, comprised depression (n=639), generalized anxiety disorder (n=310), schizophrenia (n=133), and bipolar disorder (n=93). The public repository provided both raw optical intensity data and preprocessed cerebral hemodynamic signals, accompanied by demographic metadata and standardized diagnostic labels. This dataset was uniquely positioned to support diverse research objectives, including transdiagnostic comparative analyses of functional neural patterns, the robust training of machine learning and deep learning models for multiclass psychiatric classification, and the benchmarking of fNIRS signal processing pipelines.