|国家科技期刊平台
首页|期刊导航|中国康复理论与实践|近20年国际人工智能赋能特殊儿童诊断及干预研究的可视化分析

近20年国际人工智能赋能特殊儿童诊断及干预研究的可视化分析OA北大核心CSTPCD

International researches on artificial intelligence enabled diagnosis and intervention for children with disabilities in the past two decades:a visualized analysis

中文摘要英文摘要

目的 了解近20年国际人工智能赋能特殊儿童临床诊断、治疗干预的研究现状、研究热点与演进路径,并预测未来的研究趋势. 方法 检索2004年至2023年Web of Science数据库核心合集中关于人工智能赋能特殊儿童诊断和干预研究的相关文献.采用CiteSpace 6.3.R1对其发文量、国家/地区、机构、作者、共现关键词、关键词聚类和时区图谱进行可视化分析. 结果 纳入314篇文献.欧美国家的高等院校主导人工智能赋能特殊儿童诊断和干预研究,Journal of Autism and Developmental Disorders是发文量最多的期刊,美国是发文量最多的国家,发文量最多的是以Acharya U Rajendra为代表的研究团队,该团队主要贡献是建立数据测试和孤独症患者的诊断模型.关键词主要为人工智能、孤独症谱系障碍、儿童、机器学习、深度学习、分类、诊断、青少年、个性化、识别等,关键词LLR聚类分析得到10个聚类集群.诊断、孤独症谱系障碍、注意力缺陷多动障碍、深度学习、机器学习的相关研究为研究热点.研究对象主要包括孤独症谱系障碍、注意力缺陷多动、学习障碍、Down综合征、视力障碍、脑瘫、智力障碍.人工智能技术主要包括深度神经网络、智能手表、面部表情识别、虚拟现实、机器人、机器学习、辅助技术、智能眼镜等.应用场景主要包括诊断与筛查、治疗、社会沟通、注意力训练、导航和物体识别、运动分析和治疗、认知治疗等.该领域的演化路径及发展趋势表现为:特殊儿童的诊断和干预服务,从最初的脑瘫和学习障碍,扩展到多种类型;从最初的分类和诊断,发展到多方面的识别和干预;从最初的传统技术,引入了新兴技术,为个性化和精准化教育提供更多的技术支持和方法创新,形成了多技术结合的干预模式. 结论 人工智能技术在特殊儿童诊断及干预的研究领域呈现显著上升趋势.研究重点逐渐从基础的分类和诊断转向更为复杂的诊断与个性化干预,尤其是在孤独症谱系障碍和注意力缺陷多动障碍领域.深度学习和机器学习等前沿技术的应用,正在推动该领域向更精准、更个性化的方向发展,多技术结合的干预模式成为该领域的未来发展趋势.

Objective To analyze the current status,hot topics and evolution paths of international research on artificial intelligence(AI)enabled clinical diagnosis and intervention for special children in the past 20 years,and to predict the future research trends. Methods The relevant literature of AI enabled diagnosis and intervention for special children in the Web of Science Core Collection(WoSCC)database from 2004 to 2023 was retrieved.CiteSpace 6.3.R1 was used to visualize the publi-cation volume,countries(regions),institutions,authors,co-occurrence keywords,keyword clustering and time zone map of international researches. Results A total of 314 articles were included.Famous universities in Europe and America dominated the research on AI enabled diagnosis and intervention for special children.Journal of Autism and Developmental Disorders was the journal with the most publications.The United States was the country with the most publications.The research team with the most publications was led by Acharya U Rajendra,whose main contribution was to establish data testing and diagnosis models for autism spectrum disorder(ASD)patients.The main keywords were artificial in-telligence,ASD,children,machine learning,deep learning,classification,diagnosis,adolescents,individual,rec-ognition,etc.The LLR keyword clustering analysis resulted in ten clusters.The research hotspots were diagnosis,ASD,attention deficit hyperactivity disorder,deep learning and machine learning.The research subjects mainly included ASD,attention deficit hyperactivity disorder,learning disability,Down syndrome,visual impairment,ce-rebral palsy and intellectual disability.AI technologies mainly included deep neural network,smart watch,facial expression recognition,virtual reality,robot,machine learning,assistive technology,smart glasses,etc.The appli-cation scenarios mainly included diagnosis and screening,treatment,social communication,attention training,navigation and object recognition,motion analysis and treatment,and cognitive therapy,etc.The evolution path and development trend of this field were:the diagnosis and intervention services for special children came from the initial cerebral palsy and learning disability,to multiple types;from the initial classification and diagnosis,to multiple aspects of recognition and intervention;from the initial traditional technology,to new technologies,pro-viding more technical support and method innovation for their personalized and precise education,and forming a multi-technology combination intervention mode. Conclusion The application of AI technologies in the diagnosis and intervention of children with disabilities shows a sig-nificant upward trend.The focus of research is gradually shifting from basic classification and diagnosis to more complex identification and personalized intervention,particularly in areas such as ASD and attention deficit hy-peractivity disorder.The use of cutting-edge technologies like deep learning and machine learning is propelling the field towards more precise and personalized treatment.The intervention model combining multiple technolo-gies is becoming the future trend of development in this field.

王振洲;张杨

乐山师范学院特殊教育学院,四川乐山市 614000

临床医学

人工智能残疾儿童诊断干预可视化分析

artificial intelligencechildren with disabilitiesdiagnosisinterventionvisualized analysis

《中国康复理论与实践》 2024 (004)

404-415 / 12

教育部人文社会科学研究青年基金项目(No.21XJC880006) Supported by Youth Project of Ministry of Education Foundation on Humanities and Social Science(No.21XJC880006)

10.3969/j.issn.1006-9771.2024.04.005

评论