Involvement Of The Motor System In Speech Recognition

Hearing impairment,if not identified and intervened in a timely manner,can have negative effects on development,cognition,psychosocial well-being,quality of life,and economic independence.Hearing aids can improve auditory input,but often do not restore the distorted patterns of neural activity and speech intelligibility associated with auditory loss.However,endogenous factors such as attentional state and rhythmic information regulate the activity of the sensory system just as profoundly.In the auditory system,based on motor control theory,studies have identified a large number of endogenous factors including temporal prediction information and motor nerve entrainment(NE)from the motor system,so the involvement pattern of the motor system in auditory neural activity is of high research value,and is important for finding augmentation solutions to facilitate speech recognition activities in patients with hearing loss in addition to exogenous aids such as hearing aids It is important to find a solution to facilitate speech recognition activities in patients with hearing loss in addition to external aids such as hearing aids.In the study of neural activity,magnetoencephalography has the advantage of high spatial and temporal resolution and can accurately reflect the brain activity during hearing activity.In this paper,we propose a brain dynamics model of the motor system to the auditory system during speech recognition activity.The model is based on high-precision magnetoencephalographic data,and uses phase locked value(PLV),phase amplitude coupling(PAC),and phase transfer entropy(PTE)techniques to complete the analysis of key brain regions(ROI)in speech recognition,and the analysis of brain networks between motor systems,and to complete the comparison of the differences between the brain networks in the model in passive noise listening and speech recognition tasks.The main study components as well as conclusions are as follows:1.Analysis of slow rhythmic activity in auditory cortex and phase correlation in motor cortex.The nodal connections and frequencies of synchronous activity during speech recognition activity were first determined by networked PLV,and the phase locking values between the δ and θ bands were analyzed between the frontotemporal lobe as well as the full nodal network of motor sensory cortex,and a mean difference accumulation algorithm was proposed to find the most significant phase-determined functional connection differences between the 14 nodes in both auditory modalities located in the δ band before the central return to the left transverse temporal lobe node low-frequency connections(maximum PLV of 0.588,cumulative difference of 0.187,t=0.07,p<0.05).The results showed the presence of endogenous synchronization between the motor system and the auditory system under δ oscillations during speech recognition activity.2.Asymmetric principle PAC analysis between key ROIs.Two models of neural activity for speech recognition were determined based on the asymmetry principle.PAC values in the δ band from 2 to 4 Hz and in the β and γ bands from 15 to 35 Hz were analyzed within eight nodes in sensorimotor and auditory cortex,and the nodes where the maximum PAC values appeared,and the high and low frequencies were recorded.The maximum PAC values were found to be concentrated at 2 Hz of low frequency phase oscillations.During speech recognition the left transverse temporal lobe PAC values were maximum(maximum 0.240,mean 0.182,low frequency 2 Hz,high frequency 34 Hz and20 Hz)and showed the greatest difference with passive noise hearing.The experimental results show two possible models: A.2-4Hz activity in the left anterior gyrus provides information to 2-4Hz in the left transverse temporal lobe,which is the active mode in the motor system theory.B.20 Hz and 34 Hz activity in the left transverse temporal lobe provides information to itself,while providing information to the left anterior gyrus through 2-4Hz activity,which is the passive mode.3.Directional analysis of the connections between brain regions based on phase transfer entropy.The phase transfer entropy PTE was used to perform directional detection of each frequency oscillation between the frontotemporal lobe as well as the full nodal network of motor sensory cortex to determine the direction of information flow during auditory activity.Under the δ-band,this paper reports four sets of left precordial phase shifts reversed in speech recognition,and the largest set associated with motor cortex is the reversal of the right posterior back to the left precordial,which changes from the right posterior back to the left precordial during speech recognition(maximum PTE value 0.523);under the β-band,this paper reports two sets of reversals,but with lower intensity(mean value 0.52,maximum 0.55).The results suggest a transfer of information from left motor cortex to primary auditory cortex under δ-band during speech recognition,but at the same time no directionality was found under β-band,indicating that the motor system does not provide temporal information under speech recognition.Thus the present study supports the theory of motor control of the motor system in auditory activity regarding top-down prediction of speech content and modulation of auditory cortical activity,corresponding to the active model in PAC analysis.Also this study emphasizes the role of δ oscillations in information flow activity.