דף הבית » BCN20000: DERMOSCOPIC LESIONS IN THE WILD
A PREPRINTMarc Combalia1, Noel C. F. Codella2, Veronica Rotemberg3, Brian Helba4, Veronica Vilaplana5, Ofer Reiter3, CristinaCarrera1, Alicia Barreiro1, Allan C. Halpern3, Susana Puig1, and Josep Malvehy1ABSTRACTThis article summarizes the BCN20000 dataset, composed of 19424 dermoscopic images of skinlesions captured from 2010 to 2016 in the facilities of the Hospital Clínic in Barcelona. With thisdataset, we aim to study the problem of unconstrained classification of dermoscopic images of skincancer, including lesions found in hard-to-diagnose locations (nails and mucosa), large lesions whichdo not fit in the aperture of the dermoscopy device, and hypo-pigmented lesions. The BCN20000will be provided to the participants of the ISIC Challenge 2019 [8], where they will be asked to trainalgorithms to classify dermoscopic images of skin cancer automatically
