Feature Study Of Hysterical Blindness EEG Based On Combined-channel Information

The studied hysterical blindness in this paper is one typical symptom among lots of various presentations related to hysteria. On a large extent, the diagnosis is still made mainly by experience and judgment of both symptoms and description of the patients. The evidence obtained from the specific data analysis on relationship between hysteria and brain function remains few. And feature study of hysteria electroencephalograph (EEG) appears to be even rarer. The existed studies of hysteria diagnosis need a high cost, and results in a major economic burden for the patients. An appropriate feature study of hysteria EEG would not only provide scientific explanation to neural mechanisms of the disease, but also make improvements in the diagnosis and management. The patients may also get benefit from it for their better treatment and recovery, as well as the reduction of their economic burden.In this paper, the background and development of hysteria study was summarized. The hysterical blindness EEG was presented and the feature in time domain was analyzed. Normal EEG signals collected from the adjacent positioned electrodes show similar waves, and the symmetrical parts of the brain have the same function. Based on this feature, an idea for the novel feature extraction for hysterical blindness EEGwas employed in this paper which utilized combined-channel information. After channels had been combined, the sliding-window-FastICA was respectively applied to process the combined normal EEG and the combined hysteria EEG. The kurtosis features were calculated from the processed signals. According to the feature analysis results, a brain dysfunction region was located at the occipital lobe with 01 and 02 electrode. Furthermore, new abnormality region was found at the parietal lobe with C3, C4, P3, and P4 electrode in our case that gave us a new perspective for understanding hysterical blindness. The relationship analysis between the hysterical blindness and the corresponding brain function was achieved through the kurtosis feature results and the brain function theory. Thus a reasonable explanation was given for the mechanisms of how the specific hysterical blindness correlated to brain function.As the comparison feature, the power spectral density (PSD) and the approximate entropy (ApEn) of normal and hysteria EEG were computed. The hysteria PSD of 01,02, P3 and P4 was up to 1385,1164.5,920.20 and 726.09 respectively. But obvious abnormality did not appear in C3 and C4. The normal ApEn of 01 and 02 were in the range of [0.731,0.881] and [0.674,0.827]. The hysteria ApEn of 01 and 02 distributed in [0.479,0.603] and [0.450,0.573]. The contrast showed the distinct inhibition in 01 and 02 under hysteria. But abnormal features did not detected in P3, P4, C3 and C4. Indicated by the PSD and ApEn results, kurtosis feature could characterize the hysteria EEG properly and effectively. Finally, the kurtosis feature of hysterical blindness EEG was computed after the cross channels were combined. The results could not help to distinguish the difference between hysteria EEG and normal EEG. Different combined patterns were compared to make the analysis complete. The study provided a new idea on analyzing the relationship between hysterical blindness and brain function. The kurtosis results were consistent with brain function and the clinical diagnosis. Our method was proved to be a useful tool to capture features in hysterical blindness EEG.