MedPMC: A Systematic Framework for Scaling High-Fidelity Medical Multimodal Data for Foundation Models

2026-07-08

AI Summary

This work addresses the scarcity of high-quality clinical image–text pairs, a key bottleneck in developing medical multimodal foundation models, as existing resources like PubMed Central often lack fidelity and clinical relevance. The authors propose MedPMC, the first automated and continuously updatable framework that extracts clinically validated image–text pairs from openly licensed literature through a multi-stage pipeline involving figure panel detection, image separation, caption alignment, and medical taxonomy classification. Using this approach, they construct a high-fidelity dataset of 11 million image–text pairs, which substantially enhances model performance: average zero-shot AUC improves by 7.1% across 26 benchmarks, visual question answering accuracy increases by up to 16.9%, and dermatology retrieval Recall@5 rises by 11.7%. The complete framework and dataset are publicly released.

Abstract

Medicine is inherently multimodal, requiring clinicians to synthesize information across diverse data streams. Yet the development of multimodal foundation models is constrained by limited access to large-scale, high-quality clinical data. Although PubMed Central (PMC) offers a complementary source of expert-authored image-text data, existing PMC-derived resources remain limited in fidelity, reproducibility, and clinical validation. We introduce MedPMC, an automated, continuously updatable framework that transforms permissively licensed literature into high-fidelity infrastructure for medical multimodal models. Applied to 6.1 million PMC articles, MedPMC curated 11 million medical image-text pairs. Component evaluations showed strong performance for initial screening (F1 = 93.2), multi-panel figure detection (F1 = 96.5), figure separation (mAP = 89.8), caption separation and alignment (F1 = 81.4; ROUGE-L = 85.3), and medical figure classification (F1 = 96.5). Manual review by five annotators, three with medical training, found 95.3% of MedPMC images medically relevant, versus 19.7% in a prior PMC-derived dataset. Across 26 benchmarks spanning 11 specialties, a MedPMC-trained CLIP-style model improved average zero-shot AUC by 7.1 percentage points over the strongest architecture-matched biomedical CLIP baseline despite using fewer than half as many image-text pairs. As the vision encoder in a multimodal large language model, it improved medical visual question-answering by 1.9 and 16.9 percentage points across two benchmarks. In 10,524 Yale New Haven Health System dermatology photographs, it improved morphology-to-image retrieval Recall@5 by 11.7 percentage points. These findings show that high-fidelity literature curation strengthens medical multimodal foundation models across benchmark and clinical settings. We publicly release the framework, corpus, benchmarks, and pretrained models.

Problem

• multimodal foundation models
• high-fidelity medical data
• PubMed Central
• medical image-text pairs
• clinical data scarcity

Innovation

• medical multimodal foundation models
• high-fidelity data curation
• PubMed Central
• automated figure-caption alignment
• clinical validation