Deep learning for biomedical texture image analysis
Vincent Andrearczyk, Paul F. Whelan · Dublin City University Open Access Institutional Repository (Dublin City University) · 2017
This paper shows promising results in the application of Convolutional Neural Networks (CNN) to biomedical imaging. Texture is often dominant in biomedical imaging and its analysis is essential to automatically obtain meaningful information. Therefore, we introduce a method using a Texture CNN for the classification of biomedical images. We test our approach on three datasets of liver tissues images and significantly improve the state of the art.