PRIMA: Pre-training with Risk-integrated Image-Metadata Alignment for Medical Diagnosis via LLM
2026-02-26 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors created PRIMA, a new method to help computers diagnose diseases by combining pictures and medical text more effectively. They improved how medical text is understood by adding expert knowledge about disease risks. Then, they trained the system to connect pictures and text better using special techniques. Their approach makes diagnosis more accurate and reliable without needing huge amounts of data or computing power. Experiments show it works better than previous methods.
multi-modal representation learningClinical ModernBERTRetrieval-Augmented Generation (RAG)dual-encoder pre-trainingDINOv3soft labelsQwen-3disease classificationsemantic alignment
Authors
Yiqing Wang, Chunming He, Ming-Chen Lu, Mercy Pawar, Leslie Niziol, Maria Woodward, Sina Farsiu
Abstract
Medical diagnosis requires the effective synthesis of visual manifestations and clinical metadata. However, existing methods often treat metadata as isolated tags, failing to exploit the rich semantic knowledge embedded in clinical descriptions. We propose PRIMA (Pre-training with Risk-integrated Image-Metadata Alignment), a framework that integrates domain-specific knowledge into multi-modal representation learning. We first curate an expert corpus of risk-disease correlations via Retrieval-Augmented Generation (RAG) to refine Clinical ModernBERT, embedding diagnostic priors into the text encoder. To bridge the modality gap, we introduce a dual-encoder pre-training strategy utilizing DINOv3 and our refined BERT, optimized by a suite of four complementary loss functions. These losses are designed to capture multi-granular semantic alignment and handle the ambiguity of clinical correlations through soft labels. Finally, we leverage Qwen-3 to fuse these aligned features for precise disease classification. Extensive experiments demonstrate that PRIMA effectively harmonizes pixel-level features with abstract clinical expertise, significantly outperforming other state-of-the-art methods. Notably, our framework achieves superior robustness without the need for massive data collection or exhaustive computational resources. Our code will be made public upon acceptance.