Abstract
Seborrheic keratosis (SK) and flat warts (FW) are two common occurrences of skin disease, both of which are clinically difficult to identify. However, the treatment and prognosis of the two diseases are very different, so whether they can be accurately distinguished by dermatologists is significant and important in clinical, and it helps to provide reliable and accurate decision-making for better treatment. This paper presents a deep convolution neural network discriminator for distinguishing SK and FW. The SK and FW discriminator (SFD) aims to identify and diagnose the confocal laser scanning microscope images of SK and FW by deep convolution neural network. The experimental results show that SFD performs almost equally well compared with individual dermatologists, the thediscriminator can be used to identify and diagnose between SK and FW.