Intelligent Classification and Retrieval by the Interpretation of Image Features

Abstract

The development of computer-aided diagnosis has undergone a significant transformation from multiple aspects. In the past, it was only used by radiology departments, but now it has expanded to almost every clinical specialty's imaging practices. It influences clinical diagnosis by analyzing the relationship between images and diseases, as well as their classification and retrieval. Deep learning-based convolutional neural networks or vision transformers are also leading the development in this field. This chapter introduces relevant topics on image recognition for colonoscopy and colorectal cancer, including the introduction of several articles including the early estimation of cancer by melanosis coli, the classification of polyp, and the presentation using content-based image retrieval.