| With the development of the times,paralysis caused by stroke and spinal cord injury has shown a trend of younger people,and paralysis has also become a problem in the medical field due to its long treatment time and slow results.The methods currently being studied include acupuncture and massage therapy,neurodevelopmental therapy,drug therapy,physical therapy and so on.The research team where the subject is located combines biomedicine and electronic information disciplines,puts forward the "microelectronic myoelectric bridge" theory,and develops a limb movement reconstruction system to help patients recover.This article is responsible for the research and design of the EMG detection system.The specific work is as follows:1.The project conducts theoretical research on "microelectronic myoelectric bridge" and myoelectric signal,designed the basic framework of the detection system,and completed the concrete realization of the two parts of hardware and software.System hardware includes detection circuit module,main control circuit module and communication module,and system software includes main control circuit software and Android software.2.The project researches the theory of compressed sensing and deep learning,including the principle of compressed sensing,the two preconditions of compressed sensing,and two major types of reconstruction algorithms,which is greedy algorithm and convex optimization algorithm.3.The project used compressed sensing and deep learning algorithms to sample and reconstruct the EMG signal.First,it was verified that the EMG signal satisfies the sparsity condition of compressed sensing,and then the OMP,IHT,FPC and GPSR algorithm in the compressed sensing reconstruction algorithm were simulated,and then the reconstruction algorithm based on deep learning was simulated.By comparing the reconstruction error and reconstruction time,the reconstruction algorithm based on deep learning has a reconstruction error of 10-9 magnitude,and the reconstruction time is 0.53 s.4.The project researched the hardware part of the system,including the detection circuit module,the main control circuit module and the communication module.The project realized the software part of the system.The main control circuit module software is implemented in the Keil uVersion 5.0 integrated development environment to complete the signal preprocessing,compressed sampling and data transmission functions.The Android mobile phone software is implemented in the Android Studio integrated development environment to complete WiFi communication,signal reconstruction and UI Display and other functions. |
