Deep neural network for whole slide vein segmentation

Abstract

Semantic segmentation of medical images is an area of active research all over the world. It can dramatically improve accuracy and efficiency of diagnosis if used properly. High reliability of potential solutions is required to support specialists. In this work we introduce a novel solution to perform pixelwise segmentation of vein preparations dyed with movat stain. Our proposed deep convolutional neural network achieves the accuracy of 89 %.