关键词: 内窥镜图像, 人选特征, 卷积神经网络, 计算机辅助诊断

Abstract: As an important medical diagnostic tool, endoscope is widely used in the diagnosis and screening of various diseases. With the wide application of electronic endoscope, computer-aided diagnosis algorithms based on image processing technology are constantly emerging. This paper presents a review over the application of computer aided diagnosis of medical endoscopic images by summarizing the endoscopic image analysis methods based on hand-crafted features and convolutional neural networks respectively. Finally the advantages and disadvantages of the two methods are analyzed.

Key words: endoscope image, hand-crafted feature, convolutional neural networks, computer-aided diagnosis