Medical Image Classification with Convolutional Neural Network

Abstract

Image patch categorization is a critical job in a wide range of medical imaging applications. We created a customized Convolutional Neural Network (CNN) with a shallow convolution layer to categorize lung image patches with interstitial lung disease in this study (ILD). Despite the fact that numerous feature descriptors have been developed in recent years, they can be extremely complex and domain-specific. Our customized CNN framework, on the other hand, can learn the intrinsic image characteristics from lung image patches that are most appropriate for classification automatically and effectively. The same architecture may be used to classify medical images or textures in various ways.