#clinical-notes

#clinical-notes

Open-source NLP systems effectively identify social support and social isolation in psychiatric notes using rule-based and LLM approaches.

Younger and more diverse patient demographics were observed at MSHS compared to WCM.

The Rule-Based System (RBS) showed superior performance over LLM in classifying social support and isolation.

The study presents a novel approach to identifying social support and isolation in clinical notes using rule-based and LLM systems.

NLP systems can effectively identify social support and isolation in psychiatric clinical notes.

Rule-based systems may outperform LLMs in clinical data categorization tasks.

Creating a gold standard data set for NLP requires diverse patient notes with targeted annotations to enhance accuracy in understanding social support and isolation.