基于深度学习的OCT医学图像研究进展

近年来,深度学习的研究和应用价值在学术领域和工业领域中扮演着越来越重要的角色,并且在生物医学领域也开始崭露头角。光学相干断层扫描成像(Optical Coherence Tomography,OCT)技术具有无损和分辨率高的特点,在临床诊疗与科学研究中得到了广泛的应用。本文首先介绍了深度学习和OCT技术,然后阐释了深度学习的卷积神经网络和OCT成像的基本机制,最后概述了基于深度学习技术的OCT医学图像研究,主要包括卷积神经网络在眼科、泌尿科、心血管科和肿瘤科等相关医学研究中的应用。

Progress on OCT Medical Image Research Based on Deep Learning

In recent years,the research and application value of deep learning has played an increasingly important role in academic and industrial fields,and deep learning has now begun to emerge in the field of biomedicine.Being non-invasive,optical coherence tomography imaging technology also features high resolution.OCT technology has been widely used in clinical diagnosis and scientific research.In this paper,the knowledge of deep learning and OCT technology was firstly introduced,and then the convolutional neural network and the basic principles of OCT imaging were explained.Finally,the OCT image research based on deep learning technology was summarized,focusing mainly on convolutional neural network’s application in ophthalmology,urology,cardiology and oncology.