| Stroke is the first and key factor to cause the death or disability in human in China.2million people almost die of stroke annually,and more than 80% of the survivors,to varying degrees,yield motor dysfunction in the upper or lower limbs.These weaken or lose the daily behavior ability for the patients,and also causes some tremendous burden on the economics,families and societies.Accurate and intelligent treatment has become a challenging topic for stroke rehabilitation.To build and implement an effective rehabilitation strategy,it is the key to develop some reasonable technologies and methodologies for assessing the rehabilitation effectiveness.The lesion of the cerebral vascular after stroke causes the brain structural and functional disorders,which leads to the abnormal activation over the electrophysiological activity and the changes of information interaction among the brain regions.Therefore,it is necessary and meaningful to deeply explore the potential mechanism of functional cortical coupling(FCCC)in stroke.This study will be of both theoretical value and clinical significance for the rehabilitation evaluation and pathological mechanism after stroke.This study mainly focused on the information transmission and synergistic interaction among cerebral cortex regions for stroke rehabilitation,from the perspective of both the neural mass coupling calculation model and functional corticocortical coupling characteristics.Firstly,multi dynamic single channel and coupled two channels models were built on basis of the internal physiological structure of neural mass.Then the key indexes of motion control were obtained by analyzing the model parameters based on different EEG rhythms and coupling characteristics.Secondly,multi-channel EEG synchronous coupling methods,Gabor wavelet transfer entropy and multiscale multivariable transfer entropy,were proposed to explore the local frequency band and multiscale coupling characteristics of FCCC from the views of the directionality,rhythm and multiscale.Thirdly,the correlation between the multi-level coupling characteristic indexes and the clinical evaluation scale were analyzed.These researches screened some quantitative evaluation indexes and excavated the FCCC mechanism in the motor control.Above all,the innovation of this study mainly consisted of the following aspects:(1)To explore the rhythmic oscillation of the cortical regions based on the working mode of multi-level neuron group cooperation and interaction,the multi dynamic single channel and coupled two channel models were structured to realize the signal simulation with rhythm and coupling characteristics.Furthermore,we explored the influence of model parameters on the output signals and screen out the key parameters related to motor control.It is of great significance to explore the rhythmic oscillation characteristics and the coupling mechanisms of FCCC.(2)To analyze thoroughly the local frequency band characteristics of information transmission between brain regions caused by the rhythmic change of cerebral cortical network system,a novel causality method,named Gabor Transfer Entropy(GTE),was proposed to describe the local frequency band characteristics of FCCC from the view of both nonlinearity and directionality.The EEG signals were decomposed into different frequency bands by Gabor wavelet transform.And the TE values in different frequency bands were further calculated to quantify the information transfer in two channel signals with different frequency bands.Additionally,based on the numerical simulation and neural mass models,the GTE method was proved to be effective in describing the local band coupling characteristics.(3)To explore the complex cross-scale characteristics of the FCCC and the obvious nonlinear dynamics characteristics at high spatio-temporal scales of the information interaction,the multiscale multivariate transfer entropy(MSMVTE)method was proposed to describe the direct interactions between multichannel EEG signals across multiple time scale.In this method,the coarse graining processing was applied to decompose each multivariate time series to different time scales.Then,the multivariate transfer entropy analysis was used to describe the multiscale information interaction among multichannel signals.After that,the numerical multiscale model and Chaotic model were constructed to verify the effectiveness of the MSMVTE method.(4)To investigate the nonlinear,directional,rhythm and multiscale characteristics of the information interaction between cortical brain regions for stroke patients,this study applied the GTE and MSMVTE methods to describe the local frequency-band and multiscale characteristics for stroke patients and healthy controls.Furthermore,the Pearson correlation coefficients between the FCCC characteristics and the STM assessment scale for stroke patients were calculated to explore the change trends of the FCCC indexes with high correlation with the STM scale.These researches confirmed the availability of the FCCC to describe the rhythm characteristics and coupled oscillation mechanism of the cortical regions related to motor.This study also provided a theoretical foundation for application in the analysis of pathological mechanism and the assessment of motor function status following stroke. |
