The Location And Visualization Of Brain Injury Based On EEG Analysis Technology

The structure of the brain was symmetrical and function was contralateral. Theelectroencephalogram (EEG) that represented cerebral function states wassymmetrical too, the location of brain injury could be determined by processing andanalyzing EEG data of patients with brain injury based that. According to the data,brain electrical activity mapping could be drawn and the location of brain injury couldbe visualized. The method determined the location of brain injury was presented inthis article, it was based on the analysis of features for symmetric lead EEG data. Themethod had characteristics such as high resolution, no radiation damage, low cost,easy operation and suitable for clinical examination. It could be used to assistexamination, had a couple of advantages than traditional image technology and had acertain clinical value. In this paper, the main work content was as follows:(1) The data collection and preprocessing. Through cooperation with Wu jingHospital in Hangzhou, the EEG data of45patients with brain injury who were as casegroup and10normal people who were as control group was collected.(2)Determining the location of brain injury by analyzing features for symmetricchannels EEG data. According to the cumulative analysis research, nonlinear featureapproximate entropy (ApEn) and linear feature slow wave coefficient (SWC) werechosen as the features. The standards determined the location of brain injury was therange of the features for Symmetrical-channels of patients with brain injury and testedby random sample, CT image. Results showed that: the accuracy located by ApEn inthe restingstate and call one’s name state was89.68%,91.86%and90.68%,92.70%bySWC.(3)The study for visualization techniques of the location of brain injury. Thetopographic maps of linear features average power spectral density and nonlinearfeatures such as sample entropy(SampEn), approximate entropy(ApEn),averagepower spectral density, Permutation entropy(PmEn), Complexity C0, ComplexityLZC will be drawn by the moving average method and compared with the CT images.The visual representation effect could be compared by calculating accuracy, falsepositive and false negative. Results indicate that:1、The range of the visual effectfrom best to worst is sample entropy(SampEn), approximate entropy(ApEn),Permutation entropy(PmEn), Complexity C0, Complexity LZC;2、 the visualrepresentation effect of nonlinear features SampEn and ApEn was better than linearfeatures average power spectral density; The conclusion: The visual representationeffect of DOS patients by nonlinear features such as SampEn, ApEn are better thanthe linear such as average power spectral density. The range of the sample entropyvalue had a certain clinical value to determine the location of brain injury.