Font Size: a A A

Emotion Recognition Method For Stroke Patients Based On Tactile Interaction Rehabilitation Training

Posted on:2019-01-11Degree:MasterType:Thesis
Country:ChinaCandidate:G J HuangFull Text:PDF
GTID:2428330566996033Subject:Instrument Science and Technology
Abstract/Summary:PDF Full Text Request
In the traditional rehabilitation therapy,the rehabilitative physician can not only feel the active participation of the patients,but also observe the emotional changes such as "frustration","excitement" and "boredom" and so on in patients with different training tasks,and then adjust the rehabilitation training behavior timely according to the emotional change.According to emotional psychology,the emotional response of patients during training is a kind of psychological activity characterized by the subjective participation of patients.The existing robot assisted rehabilitation training based on human-computer interaction,the interaction between patients and rehabilitative robots is mainly involved in the active participation of the patients,but the "psychological" level is not taken into consideration.To solve the problem,a method of emotion recognition for stroke patients with haptic interactive rehabilitation training is studied in this paper.We take "frustration","excitement" and "boredom" as the target mood states from the perspective of patients' active "psychological" participation in rehabilitation training.The specific work is as follows:(1)The causes,characteristics and clinical manifestations of stroke is introduced in detail,and the limitations of existing rehabilitation methods are analyzed.In particular,the effects of robotassisted rehabilitation resistance training on the patient's physiological response and the current research status and development trend at home and abroad is discussed.(2)The experimental design of emotional physiology measurement for tactile interactive rehabilitation training.Firstly,the Omega.7 tactile feedback robot is used to construct an emotional physiological measurement experimental platform for tactile interactive rehabilitation training.Secondly,inducing "frustration","excitement" and "boredom" respectively by using the virtual resistance training tasks based on tactile rehabilitation technology and International Affective Picture System(IAPS).Finally,the emotional physiological survey questionnaire is designed.We select healthy subjects and stroke patients to carry out the emotional physiological measurement experiments for the tactile interactive rehabilitation training,and the physiological signal sensors are used to collect the subjects' physiological response parameters under different emotional states: ECG signals,skin electrical signals,pulse signals,breathing signals and facial electromyography signals(scowl muscle and zygomatic muscle).(3)The features extraction of emotional physiological response for tactile interactive rehabilitation training.In the physiological response parameters collected from the subjects,we extracted the features of physiological response under the three target emotions of "frustration","excitement" and "boredom",respectively.(4)The analysis of emotional physiological response characteristics for tactile interactive rehabilitation training.By using one-way ANOVA method,the differences between the physiological response characteristics induced by different difficulty emotional induced materials(virtual resistance training tasks and emotional pictures)were analyzed from the aspects of significance,difference and correlation,and the effects of resistance on the emotional physiological response characteristics were analyzed.(5)The emotion recognition method for haptic interactive rehabilitation training.The emotional physiological response characteristics of "frustration","excitement" and "boredom" are the input,and the result of the corresponding emotional questionnaire survey is the output,and the emotion classifier bases on BP neural network is designed to identify three different types of target emotions.
Keywords/Search Tags:rehabilitative robot, haptic interaction, feature extraction, psychophysiological response, emotion recognition
PDF Full Text Request
Related items