Biomed-Enriched: A Biomedical Dataset Enriched with LLMs for Pretraining and Extracting Rare and Hidden Content

TL;DR

Biomed-Enriched is a large biomedical text dataset created using LLMs for annotation and label propagation, enabling improved biomedical NLP tasks and clinical content access.

Contribution

The paper introduces a novel two-stage annotation process using LLMs to create a large, high-quality biomedical dataset with refined subsets for NLP applications.

Findings

Clinical upsampling improves model performance by ~5%.

Educational quality filtering enhances QA task accuracy by ~1%.

Curated subsets enable faster training convergence.

Abstract

We introduce Biomed-Enriched, a biomedical text dataset constructed from PubMed via a two-stage annotation process. In the first stage, a large language model annotates 400K paragraphs from PubMed scientific articles, assigning scores for their type (review, study, clinical case, other), domain (clinical, biomedical, other), and educational quality. The educational quality score (rated 1 to 5) estimates how useful a paragraph is for college-level learning. These annotations are then used to fine-tune a small language model, which propagates the labels across the full PMC-OA corpus. The resulting metadata allows us to extract refined subsets, including 2M clinical case paragraphs with over 450K high-quality ones from articles with commercial-use licenses, and to construct several variants via quality filtering and domain upsampling. Clinical text is typically difficult to access due to privacy constraints, as hospital records cannot be publicly shared. Hence, our dataset provides an alternative large-scale, openly available collection of clinical cases from PubMed, making it a valuable resource for biomedical and clinical NLP. Preliminary continual-pretraining experiments with OLMo2 suggest these curated subsets enable targeted improvements, with clinical upsampling boosting performance by ~5% on MMLU ProfMed and educational quality filtering improving MedQA and MedMCQA by ~1%. Combinations of these techniques led to faster convergence, reaching same performance with a third of training tokens, indicating potential for more efficient and effective biomedical pretraining strategies.