EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Optimizing Entity Recognition in Psychiatric Treatment Data with Large Language Models

Seyed Mohammad Bagher Hosseini, Mohammad Javad Momeni Nezhad, Mahdis Hosseini, Maryam Zolnoori

Studies in Health Technology and Informatics · 2025

DOI: 10.3233/shti250947 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Studies in Health Technology and Informatics, 2025. Verfasst von Seyed Mohammad Bagher Hosseini und 3 weiteren Autor:innen. Beteiligte Einrichtung: Columbia University Irving Medical Center, New York, NY. Hauptthemen laut MEDLINE: Natural Language Processing, Adverse Drug Reaction Reporting Systems, Electronic Health Records, Mental Disorders. Die Arbeit wurde bisher 4-mal zitiert (Crossref). Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Extracting nuanced adverse drug reactions (ADRs) from patient self-reported messages using is pivotal but challenging, particularly given HIPAA constraints. We investigate locally deployable small LLMs–Mistral-7B, Llama-3-8B, and Gemma-7B–for ADR extraction using the PsyTAR dataset of self-reported messages. We implement in-context learning, demonstration selection, and fine-tuning with QLoRA. Results show Mistral-7B excels in few-shot settings, and Fine-tuning with F1 = 86%. These approaches safeguard data privacy and offer resource-efficient solutions for healthcare organizations. Our pipeline enables real-time ADR monitoring, helping clinicians address concerns more swiftly and enhance patient outcomes. Findings underscore that smaller LLMs can be effectively used under strict data privacy constraints, allowing healthcare teams to quickly identify and address patient-reported ADRs. Ultimately, these accessible solutions bolster patient safety.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Natural Language ProcessingAdverse Drug Reaction Reporting SystemsElectronic Health RecordsMental DisordersDrug-Related Side Effects and Adverse ReactionsData MiningHumansLarge Language Models

Stichwörter der Autor:innen

Adverse drug reactionsEntity extractionLarge language modelsName entity recognitionPatient message

Autor:innen und Institutionen

  • Columbia University Irving Medical Center, New York, NY.
  • Columbia University School of Nursing, New York, NY.

Zitieren

APA

Hosseini, S. M. B., Momeni Nezhad, M. J., Hosseini, M., & Zolnoori, M. (2025). Optimizing Entity Recognition in Psychiatric Treatment Data with Large Language Models. Studies in Health Technology and Informatics. https://doi.org/10.3233/shti250947

Vancouver

Hosseini SMB, Momeni Nezhad MJ, Hosseini M, Zolnoori M. Optimizing Entity Recognition in Psychiatric Treatment Data with Large Language Models. Studies in Health Technology and Informatics. 2025. doi:10.3233/shti250947

BibTeX
@article{Hosseini2025,
  author = {Hosseini, Seyed Mohammad Bagher and Momeni Nezhad, Mohammad Javad and Hosseini, Mahdis and Zolnoori, Maryam},
  title = {Optimizing Entity Recognition in Psychiatric Treatment Data with Large Language Models},
  journal = {Studies in Health Technology and Informatics},
  year = {2025},
  doi = {10.3233/shti250947}
}
Kostenlos starten →