Research On Personalized Tongue Control Assistant Technology For Patients With Limb Disability

Physically disabled patients lose part of their mobility due to congenital or acquired diseases,old age and other factors.After disabled patients,they need help and care from others,which brings a certain burden to the family and society.Intelligent assistive technology can effectively solve the problems of early rehabilitation of patients with physical disability and autonomous life after disability.Tongue-controlled assistive technology is more effective for patients with moderate and severe disabilities caused by spinal nerve injury.Due to the differentiation of human oral and tongue movements,individualized control requirements have become a research hotspot in this direction.This thesis mainly studies the precise control method of tongue control assistive technology for patients with limb disability,and obtains a personalized tongue control strategy.The specific work is as follows:Firstly,the individualized control needs of patients with limb disability and the influencing factors of tongue control strategies are analyzed,the electrode sensor arrangement suitable for individualized tongue control needs is given,and the overall technical scheme of personalized tongue control assistance for patients with limb disability is designed.Based on the personalized tongue control assistance technology scheme,a personalized tongue control method based on random forest is proposed,and a personalized tongue control strategy prediction model based on random forest is established.First,the electrode sensor data collected by the tongue control device and the basic data such as the patient’s age and disability are obtained,and the above data are preprocessed to generate training samples.The random forest algorithm is used to classify and train the training samples,and a set of general tongue control instructions is generated to realize group model training.In order to meet the needs of individualized control,this thesis designs and implements an individualized data processing and optimization platform,which inputs individual patient tongue touch data into the group model,completes individualized training,and finally generates a tongue control instruction set that is more suitable for individual patients.The results before and after personalized training are compared.The experimental results show that the prediction accuracy of the model after personalized training exceeds 80%,which is about 20% higher than that of the group model,which is more suitable for the control needs of individual patients.Finally,the personalized data processing and optimization platform is tested,and the results show that the tongue control assistant technology method proposed in this thesis can realize one-click training and strategy management,and provide personalized tongue control strategies for limb disabled patients.At the same time,the operational technical requirements of model training and the hardware requirements of terminal equipment are reduced,giving consideration to accuracy and ease of use.