Deep learning distinguishes skin cancer subtypes from images without biopsy
Deep Learning for Biopsy-Free Subtyping of Basal Cell Carcinoma from Dermatoscopic Images
Computer Vision and Pattern RecognitionArtificial Intelligence
Summary
Basal Cell Carcinoma is the most common skin cancer, and doctors usually need a skin biopsy to figure out the exact subtype, which guides treatment. The authors explored using artificial intelligence to identify the subtype just from a single skin image taken by a special camera. They used advanced computer models called vision transformers that are good with small amounts of data. Their method was better than older models and previous human performance for detecting aggressive cancer types, potentially helping doctors decide treatment without invasive biopsies.
What this means in practice
- •For dermatology clinics: Use AI models to subtype basal cell carcinoma directly from dermatoscopic images, enabling less invasive diagnosis and tailored treatment planning.
- •For medical imaging developers: Incorporate vision transformer models trained on BCC dermatoscopic images to improve automated classification tools distinguishing aggressive lesion subtypes.
Tested on one dataset.
Authors
Alexandros Papadopoulos, Chrysa Episkopou, Ioannis Sarafis, Aimilios Lallas, Anastasios Delopoulos
Abstract
Basal Cell Carcinoma (BCC) is the most common type of skin cancer, accounting for nearly 80% of skin cancer di- agnoses. Its optimal clinical management is guided by the distinct histopathologic subtype, with aggressive variants requiring more drastic measures. In current clinical practice, subtyping relies on skin biopsies, a procedure both costly and invasive. In this paper, we conduct a preliminary investigation into using deep learning for BCC subtyping, solely from a single dermatoscopic image of the lesion. Given the limited data at our disposal, we employ pre-trained vision transformers (ViTs), a state-of-the-art family of models highly effective for challenging downstream tasks with limited labeled data. Through repeated stratified k-fold cross-validation, we demonstrate that ViTs can achieve superior performance (AUC 0.784 on a dataset of 1271 dermatoscopic images of various BCC subtypes) over standard CNN-based baselines as well as previously-reported human reader perfor- mance, on the task of differentiating aggressive BCCs from other subtype families. These initial findings highlight the potential of combining deep learning and dermatoscopy to provide a biopsy- free alternative for BCC subtyping, thus aiding in improving treatment planning and patient outcomes.