Revamping AI Models in Dermatology: Overcoming Critical Challenges for Enhanced Skin Lesion Diagnosis

TL;DR

This paper introduces the HOT model, a comprehensive AI system for dermatology that improves skin lesion diagnosis by handling out-of-distribution images, integrating multiple data types, and providing clinical decision support.

Contribution

The paper presents a novel hierarchical-out-of-distribution clinical triage model that enhances dermatology AI by addressing real-world challenges and integrating clinical and dermoscopic data.

Findings

HOT model effectively detects out-of-distribution images.

The model improves diagnostic accuracy with combined clinical and dermoscopic images.

Experiments demonstrate the model's robustness and clinical utility.

Abstract

The surge in developing deep learning models for diagnosing skin lesions through image analysis is notable, yet their clinical black faces challenges. Current dermatology AI models have limitations: limited number of possible diagnostic outputs, lack of real-world testing on uncommon skin lesions, inability to detect out-of-distribution images, and over-reliance on dermoscopic images. To address these, we present an All-In-One \textbf{H}ierarchical-\textbf{O}ut of Distribution-\textbf{C}linical Triage (HOT) model. For a clinical image, our model generates three outputs: a hierarchical prediction, an alert for out-of-distribution images, and a recommendation for dermoscopy if clinical image alone is insufficient for diagnosis. When the recommendation is pursued, it integrates both clinical and dermoscopic images to deliver final diagnosis. Extensive experiments on a representative cutaneous lesion dataset demonstrate the effectiveness and synergy of each component within our framework. Our versatile model provides valuable decision support for lesion diagnosis and sets a promising precedent for medical AI applications.