论文部分内容阅读
近十年来，人工智能技术快速发展并逐渐由学术界走向产业界，其在医疗领域的应用也逐渐深入。受技术和伦理的局限，人工智能在医疗领域更多处于辅助决策的地位。抑郁症作为一种常见的精神障碍，其发病率在全球日益增长，如何利用以深度学习为代表的人工智能技术手段实现对抑郁症的筛查和诊断，促进抑郁症早发现和及时治疗，具有十分重要的意义。我们对近几年以人工智能为手段的抑郁症辅助诊断技术进行了文献调研和总结，主要从人脸表情、语音语调、文本语义、姿态行为及多模态数据融合5个方面入手，介绍人工智能在面向患者日常行为分析的抑郁症辅助诊断方面的研究进展。“,”In the past decade, the development of artificial intelligence technology has gradually spread from academia to industry, and its application in the medical field has profoundly deepened. As the driving engine of artificial intelligence, the machine learning and deep learning， based on data and algorithms, contribute the core technology of artificial intelligence (AI), given the limitation of technology and the ethical consideration, AI can be more acceptedits role in auxiliary decision-making instead of independent diagnosis and treatment. Depression is one of the world wide popular mental condition, and its early diagnosis depression promot estimely treatment. AI technology may help to screen depression by analysis of the various characteristics: the physiological indicators, facial expressions, voice intonation, text semantics, gesture behavior, and with all above information calcucated together to establish a model of depression auxiliary diagnosis system. This paper summarizes the literature on the AI auxiliary diagnosis for depression based on different models as well as multimodal data fusion.