Generation of Clinical Skin Images with Pathology with Scarce Data

Artificial Intelligence (AI) has proven that can be a precious tool in the healthcare domain, via the automation of menial tasks and the assistance provided to healthcare providers and doctors. Dermatology is among the areas which can benefit from data-driven models, as the first step of identifying skin diseases typically consists of visual inspection (possibly followed by further analyses) and AI approaches are well-suited to classify images—if provided with sufficient training data. As such data is often scarce, we present a Machine Learning (ML) technique to generate synthetic but realistic skin images with a variety of conditions. Such a generator can be trained using very few samples and can significantly augment data set size for downstream tasks such as disease detection through image classification. We demonstrate our approach to generating synthetic images starting from a data set collected at an Italian hospital and consisting of a few hundred samples.