Abstract: Diseases in brain seriously threaten the health of human beings. However, with the quick development of computer technology, the medical images are increasingly used in medical diagnosis, such as  CT, MRI and PET-CT, which puts the technology of computer visualization to a new height. The diseases of the brain are various, and some of them even directly threaten human life. Therefore, the diagnosis of brain diseases in early stage is very important for the treatment of diseases. At present, deep learning and its advantage of automatic features extraction make the computer aided medical diagnosis become a research hotspot. This paper briefly introduces the imaging technology of human brain firstly. Then, it introduces the traditional methods of pathological brain and the methods based on deep learning. The comparisons, analysis and conclusion will be introduced afterwards. At last, the current situation and trend of the methods based on deep learning are shown.

Key words: diseases in brain, medical image, Magnetic Resonance Imaging（MRI）, computer-aided diagnosis

摘要： 脑部疾病严重威胁着人类的健康，而随着计算机技术的发展，CT、MRI、PET-CT等医学图像越来越多地被应用到医疗诊断工作当中，这将计算机可视化技术推向了一个新的高度。脑部疾病的病情各异，有的甚至直接危及生命，因此，脑疾病的早期诊断对疾病治疗具有十分重要的意义。目前，深度学习的火热，及其自动学习特征的优势，使得基于深度学习的计算机辅助医疗诊断成为了研究热点。将简单介绍病脑成像技术，并针对传统的病脑检测方法和基于深度学习的病脑检测方法进行比较、分析和总结，研究其发展现状和趋势，并针对不同的诊断方法对其检测性能等方面进行分析与比较。

关键词: 脑疾病, 医学图像, 核磁共振成像（MRI）, 计算机辅助诊断