Quantitative Analysis of Motor Synergies and Assessment of Upper-limb Motor Function for Post-stroke Rehabilitation Based on Multi-modal Data Fusion

 I really hate assessment stroke research. Because there are NO stroke rehab protocols to use once you have been assessed.  You describe a problem, but offer NO solutions.  Solutions, NOT guidelines.

Quantitative Analysis of Motor Synergies and Assessment of Upper-limb Motor Function for Post-stroke Rehabilitation Based on Multi-modal Data Fusion

Chen Wang1,2, Liang Peng1, Jingyue Li4, Zeng-Guang Hou*1,2,3, and Weiqun Wang1
1 State Key Laboratory of Management and Control for Complex Systems, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China; 2 University of Chinese Academy of Sciences, Beijing 100149, China; 3 CAS Center for Excellence in Brain Science and Intelligence Technology, Beijing 100190, China; 4 China Rehabilitation Research Center, Beijing Bo’ai Hospital, Beijing 100068, China. {wangchen2016,liang.peng,zengguang.hou,weiqun.wang}@ia.ac.cn, lijingyue87@126.com

Abstract. 

Upper limb functional assessment plays an important role in rehabilitation protocols after stroke.(Where the fuck are these protocols?) The current assessment process is labour-intensive and relies heavily on clinical experience. In order to objectively quantify the upper-limb motor impairments in post-stroke hemiparetic patients, this study proposes a novel assessment method capable of fusing kinematic data and surface electromyography (sEMG) signals. During goal-directed movements, multi-modal data were collected synchronously, and the intra-channel statistical features were served as the inputs of different single-modality classifiers. In addition, inter-channel synergies were quantified at the kinematic and muscular levels. Then, the outputs of single-modality classifiers and synergy quantification were integrated by a multi-modal fusion scheme, and three types of machine learning algorithms were tested in the assessment framework. Experimental results demonstrated the classification accuracy was improved to 94.4% by integrating the intra-channel and inter-channel characteristics from different modalities, and the assessment output exhibited a high consistency with the score of Fugl-Meyer Assessment (FMA). The promising performance suggests that the proposed method has the potential to evaluate the effectiveness of post-stroke rehabilitation.