Early identification of stroke through deep learning with multi-modal human speech and movement data

作     者：Zijun Ou Haitao Wang Bin Zhang Haobang Liang Bei Hu Longlong Ren Yanjuan Liu Yuhu Zhang Chengbo Dai Hejun Wu Weifeng Li Xin Li Zijun Ou;Haitao Wang;Bin Zhang;Haobang Liang;Bei Hu;Longlong Ren;Yanjuan Liu;Yuhu Zhang;Chengbo Dai;Hejun Wu;Weifeng Li;Xin Li

作者机构：School of Computer Science and EngineeringSun Yat-sen UniversityGuangzhouGuangdong ProvinceChina Department of NeurologyGuangdong Neuroscience InstituteGuangdong Provincial People’s Hospital(Guangdong Academy of Medical Sciences)Southern Medical UniversityGuangzhouGuangdong ProvinceChina Department of Emergency MedicineGuangdong Provincial People’s Hospital(Guangdong Academy of Medical Sciences)Southern Medical UniversityGuangzhouGuangdong ProvinceChina 

出 版 物：《Neural Regeneration Research》 (中国神经再生研究（英文版）)

年 卷 期：2025年第20卷第1期

页      面：234-241页

核心收录：

学科分类：12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 1002[医学-临床医学] 081104[工学-模式识别与智能系统] 08[工学] 0835[工学-软件工程] 100204[医学-神经病学] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 10[医学] 

基　　金：supported by the Ministry of Science and Technology of China,No.2020AAA0109605(to XL) Meizhou Major Scientific and Technological Innovation Platforms Projects of Guangdong Provincial Science & Technology Plan Projects,No.2019A0102005(to HW) 

主　　题：artificial intelligence deep learning diagnosis early detection FAST screening stroke 

摘      要：Early identification and treatment of stroke can greatly improve patient outcomes and quality of *** clinical tests such as the Cincinnati Pre-hospital Stroke Scale(CPSS)and the Face Arm Speech Test(FAST)are commonly used for stroke screening,accurate administration is dependent on specialized *** this study,we proposed a novel multimodal deep learning approach,based on the FAST,for assessing suspected stroke patients exhibiting symptoms such as limb weakness,facial paresis,and speech disorders in acute *** collected a dataset comprising videos and audio recordings of emergency room patients performing designated limb movements,facial expressions,and speech tests based on the *** compared the constructed deep learning model,which was designed to process multi-modal datasets,with six prior models that achieved good action classification performance,including the I3D,SlowFast,X3D,TPN,TimeSformer,and *** found that the findings of our deep learning model had a higher clinical value compared with the other ***,the multi-modal model outperformed its single-module variants,highlighting the benefit of utilizing multiple types of patient data,such as action videos and speech *** results indicate that a multi-modal deep learning model combined with the FAST could greatly improve the accuracy and sensitivity of early stroke identification of stroke,thus providing a practical and powerful tool for assessing stroke patients in an emergency clinical setting.