Abstract
Neuro-linguistic programming for mental health : what the future holds for clinical study
Author(s): Joydeep Sarkar*The current report has demonstrated the application of transformer based Natural Language Processing (NLP) model in identification (presence/absence) of complex symptoms of patients with Major Depressive Disorder (MDD) [1]. The model was specifically developed for identification of 3 symptoms: Anhedonia, suicidal ideation with intent or plan and suicidal ideation without intent or known plan. The NLP model was trained to clinical narratives of MDD patients obtained from progress notes at a specialty care institution. The model was able to demonstrate remarkable accuracy (both sensitivity and specificity) in the identification of said symptoms from verbose descriptions. Additionally, the report characterizes the robustness of the model to changes in the language, keywords etc.