人工智能在帕金森病构音障碍研究中的应用进展  

Application progress of artificial intelligence in the study of speech disorders in Parkinson′s disease

作　　者：韦朝霞[1] 李丽华[2] 罗庆禄 唐桂华 Wei Zhaoxia;Li Lihua;Luo Qinglu;Tang Guihua(Department of Rehabilitation,the Tenth Affiliated Hospital of Southern Medical University,Dongguan Key Specialty of Traditional Chinese Medicine(Rehabilitation Department),Dongguan Experimental Centre for Sports Rehabilitation Research,Dongguan 523058,China;Future Institute of Technology,South China Normal University,Guangzhou 510631,China)

机构地区：[1]南方医科大学第十附属医院(东莞市人民医院)康复医学科,东莞市中医药局临床重点(特色)专科,东莞市运动康复研究实验中心,东莞523058 [2]华南师范大学未来科技研究院,广州510631

出　　处：《中华神经科杂志》2024年第11期1259-1263,共5页Chinese Journal of Neurology

摘　　要：帕金森病是一种严重的多发于老年人的神经退行性疾病,对患者的发音有很大影响,可导致构音障碍。人工智能的出现为帕金森病相关构音障碍的早期发现、评估、诊断和康复开辟了新途径。文中探讨了人工智能在帕金森病构音障碍研究中的应用,讨论了所采用的方法和技术,如语音信号处理、特征提取技术和分类算法,以及它们所能检测到的症状,包括语音质量、发音、语速和前音,重点介绍了人工智能在早期诊断、疾病进展监测、治疗评估和远程康复中的作用。Parkinson′s disease(PD)is a neurodegenerative disorder that significantly impacts speech and voice,leading to hypokinetic dysarthria,a motor speech disorder.The advent of artificial intelligence(AI)has opened new avenues for the assessment,diagnosis,and rehabilitation of PD-associated speech impairments.This review explores the application of AI in the study of PD speech disorders,focusing on automated speech analysis,machine learning algorithms,and the development of speech pathology databases.This review also discusses the methodologies and technologies employed,such as speech signal processing,feature extraction techniques,classification algorithms,and the symptoms they can detect,including voice quality,articulation,speech rate,and prosody.The role of AI in early diagnosis,disease progression monitoring,treatment evaluation,and remote rehabilitation is highlighted.The review concludes with a discussion on the potential and challenges of AI in this field and recommendations for future research.

关 键 词：帕金森病 构音障碍 人工智能 语音分析 机器学习 

分 类 号：R742.5[医药卫生—神经病学与精神病学]