A Global Atlas of Digital Dermatology to Map Innovation and Disparities

2025-12-27

medRxiv

AI Summary

Current digital dermatology datasets lack quantitative assessment of clinical coverage breadth and novelty, leading to significant biases in AI models across populations and disease spectra. This work proposes SkinMap, a multimodal framework that integrates over 1.1 million publicly available skin images to construct a queryable semantic atlas, enabling the first panoramic audit of the global dermatological data foundation. Through multimodal semantic mapping and coverage quantification, the study systematically evaluates dataset novelty, redundancy, and representational gaps—particularly the underrepresentation of darker-skinned individuals (5.8%) and children (3.0%), as well as rare disease phenotypes. The analysis reveals that increased data volume does not consistently yield additional information gain and provides actionable guidance for strategic data collection.

Technology Category

• Data Mining & Knowledge Management: Mining of Visual, Multimedia & Multimodal Data
• Computer Vision: Multi-modal Vision
• Machine Learning: Multimodal Learning

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Security and Privacy: Data transparency and provenance
• Web Mining and Content Analysis: Web data quality in the era of algorithmically-generated content

Abstract

The adoption of artificial intelligence in dermatology promises democratized access to healthcare, but model reliability depends on the quality and comprehensiveness of the data fueling these models. Despite rapid growth in publicly available dermatology images, the field lacks quantitative key performance indicators to measure whether new datasets expand clinical coverage or merely replicate what is already known. Here we present SkinMap, a multi-modal framework for the first comprehensive audit of the field's entire data basis. We unify the publicly available dermatology datasets into a single, queryable semantic atlas comprising more than 1.1 million images of skin conditions and quantify (i) informational novelty over time, (ii) dataset redundancy, and (iii) representation gaps across demographics and diagnoses. Despite exponential growth in dataset sizes, informational novelty across time has somewhat plateaued: Some clusters, such as common neoplasms on fair skin, are densely populated, while underrepresented skin types and many rare diseases remain unaddressed. We further identify structural gaps in coverage: Darker skin tones (Fitzpatrick V-VI) constitute only 5.8% of images and pediatric patients only 3.0%, while many rare diseases and phenotype combinations remain sparsely represented. SkinMap provides infrastructure to measure blind spots and steer strategic data acquisition toward undercovered regions of clinical space.

Problem

• digital dermatology
• dataset bias
• representation gaps
• clinical coverage
• data diversity

Innovation

• SkinMap
• multimodal framework
• data audit
• representation gaps
• semantic atlas