| Although depression can seriously impair the patient’s physical and mental health,there is currently no theory that can completely explain the pathology of depression and there is no specific,quantifiable standard for the diagnosis of depression.In this context,this paper uses the complexity of nonlinear dynamics in the study of depression.Magnetoencephalography signal has some advantages,including high spatio-temporal resolution,non-invasive and easy to detect on the human body,can directly reflect the brain activity of the brain,the measurement results are not affected by the outer layer of the brain tissue,etc.,is very suitable for the study of depression.In this paper,the algorithm of complexity is used to analyze the depression magnetoencephalography under emotional stimuli to explore the related pathological changes of depression and the difference of the complexity of magnetoencephalography in patients with depression and normal.The contents of the paper can be divided into the following sections:(1)Magnetoencephalography analysis of depression based on approximate entropyIn this part,the parameters of the approximate entropy algorithm suitable for magnetoencephalography are selected first by experiments,then the approximate entropy is applied to the magnetoencephalography of the depression group and the normal control group,and the approximate entropy of each group was calculated under the positive,neutral and negative emotional stimuli respectively.It is found that under three emotional stimuli,the approximate entropy of magnetoencephalography in depressive patients group is lower than that in normal group in most of the channels,the approximate entropy of magnetoencephalography in depressive patients group is lower than normal group in all brain regions,and the difference between two groups of approximate entropy is the most obvious in the frontal area.(2)Magnetoencephalography analysis of depression based on sample entropySince the approximate entropy algorithm will perform self-matching when calculating the complexity,it may result in the deviation of the results.In this section,the sample entropy is used in the analysis of magnetoencephalography of the depressive patient group and the normal human control group,the corresponding sample entropy under positive,neutral and negative emotional stimuli is calculated respectively.The experimental results obtained are basically the same as those obtained by the approximate entropy experiment.However,the entropy obtained by the sample entropy algorithm is larger than the approximate entropy algorithm,and the relative difference of the complexity of the two data sets obtained by the sample entropy is more significant than the approximate entropy.And it is found that the sample entropy is more suitable than the approximate entropy to distinguish the magnetoencephalography of normal people and depressive patients under emotional stimulation.(3)Complexity analysis system based on java webThis part implements a complexity analysis system using java web related technologies.In this system,users upload data sources by uploading files.They can choose algorithm parameters for approximate entropy and sample entropy calculation.They can freely combine the source data that need to be calculated,and support the contrast experiment.Users can see the calculation progress,and can view the experimental results which have been calculated,including approximate entropy,sample entropy,and relative differences.These results are stored in the current account and can be viewed at any time.This system uses a visual interface to guide the user to operate,and the users can easily calculate the approximate entropy and sample entropy of the data source. |
