Objective Assessment Of Brain Cognitive Load Based On Morphological Features Of Pulse Wave

Cognitive load refers to the total amount of cognitive resources consumed by a person due to information processing in the brain in the process of learning knowledge or solving problems.Moderate cognitive load can maximize the task completion efficiency.If the cognitive load is too heavy,it will lead to a decrease in the efficiency of task completion.Therefore,we need timely assess cognitive load to control it within the optimal load range of an individual.Pulse wave has the advantages of "easy to collect" and "easy to analyze".Its morphological features contain a lot of information about the activity of the autonomic nervous system related to cognitive load,so it can realize the objective assessment of cognitive load.This paper proposes an objective cognitive load assessment method based on the morphological features of pulse wave.The method realizes two classification tasks according to actual needs: one is the rough classification of two cognitive load states induced by tasks of different difficulty levels;the other is the fine classification of "high-load match level" and "high-load overpressure level" under high-load tasks.The main contributions and research contents of this paper are as follows:(1)About signal preprocessing and feature extraction,the periodic screening algorithm is used to exclude abnormal heart-beat intervals,to solve the problem that the pulse wave signal is susceptible to motion artifacts.The backward and forward traversal of the maximum slope point and the difference method are used to locate the feature points of pulse wave,to improve the stability of feature points identification.Based on the extracted feature points and the regulation mechanism of autonomic nervous system to cardiovascular,a set of multidimensional pulse wave morphological features extraction schemes is proposed.(2)About the identification of the two cognitive load states,this paper carries out feature debaselining aiming at the individual differences in morphological characteristics.Feature selection algorithm and classification algorithm as well as a 4-dimensional optimal feature combination were used to reach a recognition accuracy of 82%.Compared with the unprocessed features,the recognition accuracy of the debaselined features was improved by11.9%.The corresponding 4-dimensional optimal features are closely related to cardiovascular activity,suggesting that the indicators reflecting cardiovascular activity can well distinguish the two cognitive load states.Compared with the combination of features used in existing studies,the morphological features selected herein have a better identification effect under the same dimension.(3)This paper innovatively uses the multidimensional morphological features of pulse wave to reclassify the cognitive load level under high-load tasks.Through the feature significance test and feature selection algorithm,the most favorable feature combination was screened out,and the support vector machine classifier was used to realize the recognition of two cognitive load levels,reaching a recognition accuracy of 95.3%.Compared with the features proposed by existing studies,the morphological features extracted by this paper can better distinguish cognitive load levels.