DermINO: Hybrid Pretraining for a Versatile Dermatology Foundation Model

2025-08-16

AI Summary

Skin diseases affect approximately 70% of the global population, yet diagnosis remains challenging due to clinical complexity and a severe shortage of dermatologists—especially at primary care levels—hindering equitable access to dermatological care. Existing AI models are limited by heavy reliance on large-scale annotated datasets and poor generalization across diverse dermatological tasks. To address these bottlenecks, we propose the first dermatology-specific foundation model, integrating self-supervised learning, semi-supervised learning, knowledge-guided prototype initialization, and federated learning. Trained on 430,000 multi-source skin images, the model achieves state-of-the-art performance across 20 benchmark datasets: it attains 95.79% accuracy in malignant lesion classification—surpassing dermatologists’ average (73.66%)—and boosts clinician diagnostic accuracy by 17.21% under AI assistance. Moreover, it delivers leading performance in lesion segmentation, severity grading, and medical image generation, establishing a unified, generalizable framework for dermatological AI.

Technology Category

• Machine Learning: Deep Neural Architectures and Foundation Models
• Computer Vision: Diffusion Models for Vision
• Humans and AI: Other Foundations of Human Computation & AI

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization

Abstract

Skin diseases impose a substantial burden on global healthcare systems, driven by their high prevalence (affecting up to 70% of the population), complex diagnostic processes, and a critical shortage of dermatologists in resource-limited areas. While artificial intelligence(AI) tools have demonstrated promise in dermatological image analysis, current models face limitations-they often rely on large, manually labeled datasets and are built for narrow, specific tasks, making them less effective in real-world settings. To tackle these limitations, we present DermNIO, a versatile foundation model for dermatology. Trained on a curated dataset of 432,776 images from three sources (public repositories, web-sourced images, and proprietary collections), DermNIO incorporates a novel hybrid pretraining framework that augments the self-supervised learning paradigm through semi-supervised learning and knowledge-guided prototype initialization. This integrated method not only deepens the understanding of complex dermatological conditions, but also substantially enhances the generalization capability across various clinical tasks. Evaluated across 20 datasets, DermNIO consistently outperforms state-of-the-art models across a wide range of tasks. It excels in high-level clinical applications including malignancy classification, disease severity grading, multi-category diagnosis, and dermatological image caption, while also achieving state-of-the-art performance in low-level tasks such as skin lesion segmentation. Furthermore, DermNIO demonstrates strong robustness in privacy-preserving federated learning scenarios and across diverse skin types and sexes. In a blinded reader study with 23 dermatologists, DermNIO achieved 95.79% diagnostic accuracy (versus clinicians' 73.66%), and AI assistance improved clinician performance by 17.21%.

Problem

• Addresses global dermatology diagnostic challenges with AI
• Overcomes limitations of narrow-task AI models in dermatology
• Enhances generalization across clinical tasks with hybrid pretraining

Innovation

• Hybrid pretraining combines self-supervised and semi-supervised learning
• Knowledge-guided prototype initialization enhances model generalization
• Robust performance in federated learning and diverse demographics