Research On Brain Mechanism Of Abnormal Readiness Potential And Decoding Of Motor Intention

Readiness potentials are slow-rising negative potentials that appear in the centralfrontal area of the scalp before movement execution.It is a neural information that can be recorded by electrode.Researchers generally believe that the potential represent the brain’s activity for motor preparation.Motor intention is the cognitive activity of the brain before execution of movement.The process of decoding motor intention from neural information is called motor intention decoding.Readiness potential is the main neural information currently used in motor intention decoding.Decoding of motor intention plays an important role in brain-computer interface and the diagnosis of neurological diseases.Previous studies have showed that the readiness potentials of some healthy subjects disappeared,which greatly reduced the accuracy of motor intention decoding of these subjects,and brought great challenge to analysis of motor intention in practical application scenarios.Therefore,the motivations of this thesis are:determining whether the disappearance of readiness potentials is universal among the population subjects;exploring the factors and neural mechanisms leading to this phenomenon;attempting to give a solution to improve the accuracy of motor intention decoding in case of absence of the readiness potentials;applying the EEG features of motor preparation to the clinical assessment of the level of consciousness in patients with Disorder of Consciousness(DOC).The main contents and innovation points of this thesis are presented as follows:(1)The universality of the disappearance of readiness potentials was determined,and the direct cause of this phenomenon was checked.Two experimental paradigms consisting of voluntary and instructed actions task respectively were adopted,and the readiness potentials of the subjects collected in above paradigms were then analyzed with the event-related potential technique.Statistical results indicated that 48.28% of all subjects showed abnormal readiness potentials in the instructed actions task.Based on the hypothesis of selective slow cortical potential,the decrease in the proportion of negative slopes of slow potentials before movement was preliminarily determined as the direct cause of this phenomenon.This study is the first attempt to report the steady disappearance of readiness potentials.(2)The time estimation was identified as one factor leading to the disappearance of readiness potentials,and the neural mechanisms of this phenomenon was also explored.Three contrasted experiments were adopted: motor execution task,motor execution after time estimation task,and time estimation task.The EEG of the subjects were collected through above experiments,and the readiness potentials under different tasks were analyzed with event-related potential technology.The statistical results showed that the time estimation could lead to the disappearance of readiness potentials;The EEG signals were traced based on the distributed source model,and the causal network relationship between brain regions was investigated with the Adaptive Directed Transfer Function(ADTF)method,showing that the time estimation and motor preparation were performed by two distinct neural circuits,in which the Supplementary Motor Area(SMA)both played an important role.Thus,the disappearance of the readiness potential results from the counteraction of two different neural electrical activities in the Supplementary Motor Area.This study sheds light on the neural mechanisms underlying the disappearance of readiness potentials.(3)A decoding algorithm based on fusion of multi-dimensional EEG features was proposed to improve the accuracy of motor intention decoding when the readiness potentials disappear.Comparative analysis for EEG features of subjects with abnormal and normal readiness potentials during the process of motor preparation were performed,and the results showed that the time-frequency,brain network and crossfrequency coupling features during the process of motor preparation were well preserved among all above subjects.A time-frequency-cross-frequency-network features extraction and fusion method,based on EEG signals during the process of motor preparation,was proposed significantly improving the decoding accuracy when the readiness potentials disappear,and the average decoding accuracy exceeded 85%,which is better than the current decoding algorithm based solely on the readiness potentials.This study provides a new method for decoding motor intention for subjects with abnormal readiness potentials.(4)The usability of EEG features during the process of motor preparation in the assessment of the level of consciousness of DOC patients was also explored,and the EEG features that can be applied for clinical evaluation were identified.A experimental paradigm with movement task was designed for patients with DOC,and the EEG signals of patients with DOC and healthy subjects through the experiments were collected and analyzed.The results showed that readiness potentials occurred among the patients in the Minimally Conscious State(MCS)group,but the amplitudes of these potentials were not stable;The node degree of the coherent network,the global efficiency of the unbiased weighted phase lag network,and the cross-frequency coupling value showed significant differences beween the Unresponsive Wakefulness Syndrome(UWS)and Minimally Conscious State group;Combined with the brain network and cross-frequency coupling features,the classifier with support vector machine achieved a classification accuracy higher than 81% between the control group and the DOC patients group,and a classification accuracy higher than 73% between the Unresponsive Wakefulness Syndrome and the Minimally Conscious State group.This study can provide an objective basis for the assessment of the level of consciousness of patients with DOC.The work in this thesis,deepening the understanding of the brain mechanism during motor preparation,improving the performance of current movement intention decoding algorithms,and providing new ideas for clinical assessment of some brain diseases,has certain theoretical and practical values.