Research On Fast Processing Method Of 264-channel Transcranial Magnetic Stimulation Signals Based On Morpholog

Transcranial magnetic stimulation(TMS)is a non-invasive treatment technique for neurological diseases.It generates the stimulating magnetic field(SMF)through TMS coil,and then uses the SMF to affect the metabolism of human cells and other activities.Thus,TMS achieves the purpose of treating neurological diseases.Human brain cells are relatively precise and fragile,so it is necessary to detect the accuracy of the SMF before the application of TMS equipment.In the process of SMF detection,the multi-channel TMS signals representing the distribution of SMF are vulnerable to various noise interference.The traditional filtering algorithms can only filter singlechannel signal at a time,so multiple operations are required to attenuate noise in multichannel signals,which will greatly increase the processing time.In order to effectively attenuate noise in multi-channel TMS signals,the generalized morphological filtering algorithm is improved based on mathematical morphology theory.Then two novel morphological filtering algorithms are proposed,which are the adaptive weight improved generalized morphological filtering algorithm and the adaptive framing improved generalized morphological filtering algorithm.The former achieves good denoising performance through the signal grouping algorithm,the improved generalized morphological filtering algorithm and the adaptive weight algorithm,and the latter further improves the denoising performance though the adaptive framing algorithm and the improved generalized morphological filtering algorithm.The experimental results show that the two proposed algorithms have significant improvement in filtering effect and timeliness.The denoising performance of the proposed algorithms is evaluated using the signal to noise ratio,root mean square error and mean absolute error.Compared with the traditional generalized morphological filtering algorithm,the signal to noise ratio of the two proposed algorithms are increased by at least 20.63% and 32.52%,the root mean square error are reduced by at least 33.71% and 47.19%,and the mean absolute error are reduced by at least 35.21% and 49.30%.Moreover,the processing time of the two algorithms are reduced by 88.24% and 85.76%.Therefore,the two proposed algorithms can attenuate noise in multi-channel TMS signals efficiently.In order to ensure the accuracy of the SMF,it is necessary to detect the TMS signals before the application of TMS equipment.Therefore,the 7-thread host computer software is designed by the Qt cross-platform development application.The host computer connects the 264-channel high-performance TMS detection system through USB3.0 and reads the TMS signals in real time.Then the 88-channel magnetic field signals are synthesized.Finally,the monitoring interface is designed to display the 88-channel magnetic field signals in real time,and the distribution of SMF in human brain is simulated by head model and cloud image.The field experimental results show that the host computer makes great achievements in stability and visualization characteristics.It can not only detect 264-channel signals in real time,but also display the distribution of SMF in human brain,which pushes forward the development and application of TMS technology.