Summary
Basal Cell Carcinoma is a common skin cancer that needs different treatments depending on its subtype, which doctors usually find by doing a biopsy—a small tissue sample that can be painful and costly. This paper explores using an advanced type of computer model called a vision transformer to look at pictures of the skin lesions and figure out the subtype without needing a biopsy. The researchers found that their model was better at telling aggressive cancer types apart from others than traditional methods and even human doctors. While these results are preliminary, the study suggests that computers could help doctors plan treatments more easily and comfortably using just images.
Basal Cell Carcinomaskin cancerbiopsydermatoscopydeep learningvision transformersconvolutional neural networksimage classificationAUCcross-validation
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.