Maintenance Notice

Due to necessary scheduled maintenance, the JMIR Publications website will be unavailable from Wednesday, July 01, 2020 at 8:00 PM to 10:00 PM EST. We apologize in advance for any inconvenience this may cause you.

Who will be affected?

Accepted for/Published in: JMIR AI

Date Submitted: Nov 14, 2024
Date Accepted: Apr 27, 2025

The final, peer-reviewed published version of this preprint can be found here:

Leveraging Large Language Models for Accurate Retrieval of Patient Information From Medical Reports: Systematic Evaluation Study

Garcia-Carmona AM, Prieto ML, Puertas E, Beunza JJ

Leveraging Large Language Models for Accurate Retrieval of Patient Information From Medical Reports: Systematic Evaluation Study

JMIR AI 2025;4:e68776

DOI: 10.2196/68776

PMID: 40608403

PMCID: 12271962

Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.

Enhanced medical data extraction: leveraging LLMs for accurate retrieval of patient information from medical reports

  • Angel Manuel Garcia-Carmona; 
  • Maria-Lorena Prieto; 
  • Enrique Puertas; 
  • Juan-Jose Beunza

ABSTRACT

Background:

There is a need to explore the feasibility of generative solutions in extracting data from medical reports, categorized by specific criteria.

Objective:

This study proposes an advanced methodology for applying Large Language Models (LLMs) in the automated extraction of structured information from unstructured medical reports, leveraging the LangChain framework in Python.

Methods:

A comparative evaluation of state-of-the-art LLMs, including GPT-4o, LLaMA 3, LLaMA 3.1, Gemma 2, Qwen 2, and Qwen 2.5, highlights each model's unique architectural attributes and capabilities in clinical data extraction tasks. Employing zero-shot prompting techniques and embedding results in a vector database, we systematically assess the models based on metrics such as accuracy, precision, recall, and F1 score across specific data categories, including patient demographics, diagnostic information, and pharmacological details.

Results:

The results reveal high extraction efficacy, although model precision varies by data type, underscoring the challenges posed by unstructured medical text.

Conclusions:

This research demonstrates the feasibility of integrating LLMs into healthcare digitalization workflows, enabling enhanced data accessibility for clinical decision-making. Future directions emphasize optimizing model precision and generalizability through expanded training datasets and the adaptation of retrieval-augmented generation techniques to support complex, domain-specific information needs.


 Citation

Please cite as:

Garcia-Carmona AM, Prieto ML, Puertas E, Beunza JJ

Leveraging Large Language Models for Accurate Retrieval of Patient Information From Medical Reports: Systematic Evaluation Study

JMIR AI 2025;4:e68776

DOI: 10.2196/68776

PMID: 40608403

PMCID: 12271962

Download PDF


Request queued. Please wait while the file is being generated. It may take some time.

© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.