no code implementations • 15 Apr 2021 • Mohammadreza Mohseni, Jordan Yap, William Yolland, Majid Razmara, M Stella Atkins
This problem is especially important for medical image diagnosis, when an image of a hitherto unknown disease is presented for diagnosis, especially when the images come from the same image domain, such as dermoscopic skin images.
no code implementations • 22 Oct 2018 • Philipp Tschandl, Giuseppe Argenziano, Majid Razmara, Jordan Yap
Methods: A neural network was trained to predict disease classes on dermatoscopic images from three retrospectively collected image datasets containing 888, 2750 and 16691 images respectively.