Affective brain-computer interfaces (aBCIs) has important application value in the field of human-computer interaction. Electroencephalogram (EEG) has been widely concerned in the field of emotion recognition due to its advantages in time resolution, reliability and accuracy. However, the non-stationary characteristics and individual differences of EEG limit the generalization of emotion recognition model in different time and different subjects. In this paper, in order to realize the recognition of emotional states across different subjects and sessions, we proposed a new domain adaptation method, the maximum classifier difference for domain adversarial neural networks (MCD_DA). By establishing a neural network emotion recognition model, the shallow feature extractor was used to resist the domain classifier and the emotion classifier, respectively, so that the feature extractor could produce domain invariant expression, and train the decision boundary of classifier learning task specificity while realizing approximate joint distribution adaptation. The experimental results showed that the average classification accuracy of this method was 88.33% compared with 58.23% of the traditional general classifier. It improves the generalization ability of emotion brain-computer interface in practical application, and provides a new method for aBCIs to be used in practice.
Early detection and timely intervention are very essential for autism. This paper used the wavelet transform and empirical mode decomposition (EMD) to extract the features of electroencephalogram (EEG), to compare the feature differences of EEG between the autistic children and healthy children. The experimental subjects included 25 healthy children (aged 5–10 years old) and 25 children with autism (20 boys and 5 girls aged 5–10 years old) respectively. The alpha, beta, theta and delta rhythm wave spectra of the C3, C4, F3, F4, F7, F8, FP1, FP2, O1, O2, P3, P4, T3, T4, T5 and T6 channels were extracted and decomposed by EMD decomposition to obtain the intrinsic modal functions. Finally the support vector machine (SVM) classifier was used to implement assessment of autism and normal classification. The results showed that the accuracy could reach 87% and which was nearly 20% higher than that of the model combining the wavelet transform and sample entropy in the paper. Moreover, the accuracy of delta (1–4 Hz) rhythm wave was the highest among the four kinds of rhythms. And the classification accuracy of the forehead F7 channel, left FP1 channel and T6 channel in the temporal region were all up to 90%, which expressed the characteristics of EEG signals in autistic children better.
ObjectiveTo summarize the method of quality management in long term video electroencephalogram (VEEG) monitoring process.MethodsTo summarize the VEEG monitoring process in 4 935 patients, the following methods were adopted: adequate preparation before examination, selection of suitable electrode wearing methods, regular inspection of the quality of the lead wire, inspection and observation of whether the electrodes have fallen off, process inspection, behavioral intervention guidance, timely manage the artifacts, pay more attention to the inducted experimental, timely identification of paroxysmal events, standardize the procedures for the management of seizures, standardize the processing of electrode cleaning and disinfection, continuously improve the quality.ResultsFour hundred and tworoy are paroxysmal events of various types occurred during the monitoring period. All of them were handled in time and the patients were all safe. Among these events, 4 children ended the examination in ahead of the normal procedure due to fever, crying or other reasons. two patients were transferred to intensive care unit due to changes in patients ’conditions such as hypopnea and decreased oxygen saturation of artery blood of finger. The remaining 4 829 patients completed VEEG detection for 8 ~ 24 h. and got good quality images.ConclusionsQuality management is a guarantee of qualified, high quality, low artifact EEG reports.
ObjectiveTo explore the application effect of standardized management on video-electroencephalogram (VEEG) monitoring.MethodsIn January 2018, a multidisciplinary standardized management team composed with doctors, technicians, and nurses was established. The standardized management plan for VEEG monitoring from outpatient, pre-hospital appointment, hospitalization and post-discharge follow-up was developed; the special quilt for epilepsy patients was designed and customized, braided for the patient instead of shaving head, standardized the work flow of the staff, standardized the health education of the patients and their families, and standardized the quality control of the implementation process. The standardized managemen effect carried out from January to December 2018 (after standardized managemen) was compared with the management effect from January to December 2017 (before standardized managemen).ResultsAfter standardized management, the average waiting time of patients decreased from (2.08±1.13) hours to (0.53±0.21) hours, and the average hospitalization days decreased from (6.63±2.54) days to (6.14±2.17) days. The pass rate of patient preparation increased from 63.14% to 90.09%. The capture rate of seizure onset increased from 73.37% to 97.08%. The accuracy of the record increased from 33.12% to 94.10%, the doctor’s satisfaction increased from 76.34±29.53 to 97.99±9.27, and the patient’s satisfaction increased from 90.04±18.97 to 99.03±6.51. The difference was statistically significant (P<0.05).ConclusionStandardization management is conducive to ensuring the homogeneity of clinical medical care, reducing the average waiting time and the average hospitalization days, improving the capture rate and accuracy of seizures, ensuring the quality of medical care and improving patient’s satisfaction.
The effective classification of multi-task motor imagery electroencephalogram (EEG) is helpful to achieve accurate multi-dimensional human-computer interaction, and the high frequency domain specificity between subjects can improve the classification accuracy and robustness. Therefore, this paper proposed a multi-task EEG signal classification method based on adaptive time-frequency common spatial pattern (CSP) combined with convolutional neural network (CNN). The characteristics of subjects' personalized rhythm were extracted by adaptive spectrum awareness, and the spatial characteristics were calculated by using the one-versus-rest CSP, and then the composite time-domain characteristics were characterized to construct the spatial-temporal frequency multi-level fusion features. Finally, the CNN was used to perform high-precision and high-robust four-task classification. The algorithm in this paper was verified by the self-test dataset containing 10 subjects (33 ± 3 years old, inexperienced) and the dataset of the 4th 2018 Brain-Computer Interface Competition (BCI competition Ⅳ-2a). The average accuracy of the proposed algorithm for the four-task classification reached 93.96% and 84.04%, respectively. Compared with other advanced algorithms, the average classification accuracy of the proposed algorithm was significantly improved, and the accuracy range error between subjects was significantly reduced in the public dataset. The results show that the proposed algorithm has good performance in multi-task classification, and can effectively improve the classification accuracy and robustness.
To improve the performance of brain-controlled intelligent car based on motor imagery (MI), a method based on neurofeedback (NF) with electroencephalogram (EEG) for controlling intelligent car is proposed. A mental strategy of MI in which the energy column diagram of EEG features related to the mental activity is presented to subjects with visual feedback in real time to train them to quickly master the skills of MI and regulate their EEG activity, and combination of multi-features fusion of MI and multi-classifiers decision were used to control the intelligent car online. The average, maximum and minimum accuracy of identifying instructions achieved by the trained group (trained by the designed feedback system before the experiment) were 85.71%, 90.47% and 76.19%, respectively and the corresponding accuracy achieved by the control group (untrained) were 73.32%, 80.95% and 66.67%, respectively. For the trained group, the average, longest and shortest time consuming were 92 s, 101 s, and 85 s, respectively, while for the control group the corresponding time were 115.7 s, 120 s, and 110 s, respectively. According to the results described above, it is expected that this study may provide a new idea for the follow-up development of brain-controlled intelligent robot by the neurofeedback with EEG related to MI.
The directed functional connectivity in cerebral cortical is the key to understanding the pattern of the behavioral tissue. This process was studied to explore the directed functional network of rifle shooters at cerebral cortical rhythms from electroencephalogram (EEG) data, aiming to provide neurosciences basis for the future development of accelerating rifle skill learning method. The generalized orthogonalized partial directed coherence (gOPDC) algorithm was used to calculate the effective directed functional connectivity of the experts and novices in the pre-shot period. The results showed that the frontal, frontal-central, central, parietal and occipital regions were activated. Moreover, the more directed functional connections numbers in right hemispheres were observed compared to the left hemispheres. Furthermore, as compared to experts, novices had more activated regions, the stronger strength of connections and the lower value of the global efficiency during the pre-shot period. Those indirectly supported the conclusion that the novices needed to recruit more brain resources to accomplish tasks, which was consistent with " neural efficiency” hypothesis of the functional cerebral cortical in experts.
Electroencephalogram (EEG) classification for brain-computer interface (BCI) is a new way of realizing human-computer interreaction. In this paper the application of semi-supervised sparse representation classifier algorithms based on help training to EEG classification for BCI is reported. Firstly, the correlation information of the unlabeled data is obtained by sparse representation classifier and some data with high correlation selected. Secondly, the boundary information of the selected data is produced by discriminative classifier, which is the Fisher linear classifier. The final unlabeled data with high confidence are selected by a criterion containing the information of distance and direction. We applied this novel method to the three benchmark datasets, which were BCIⅠ, BCIⅡ_Ⅳ and USPS. The classification rate were 97%,82% and 84.7%, respectively. Moreover the fastest arithmetic rate was just about 0.2 s. The classification rate and efficiency results of the novel method are both better than those of S3VM and SVM, proving that the proposed method is effective.
Alzheimer's disease (AD) is the most common type of dementia and a neurodegenerative disease with progressive cognitive dysfunction as the main feature. How to identify the early changes of cognitive dysfunction and give appropriate treatments is of great significance to delay the onset of dementia. Some other researches have shown that AD is associated with abnormal changes of brain networks. To study human brain functional connectivity characteristics in AD, 16 channels electroencephalogram (EEG) were recorded under resting and eyes-closed condition in 15 AD patients and 15 subjects in the control group. The synchronization likelihood of the full-band and alpha-band (8-13 Hz) data were evaluated, which resulted in the synchronization likelihood coefficient matrices. Considering a threshold T, the matrices were converted into binary graphs. Then the graphs of two groups were measured by topological parameters including the clustering coefficient and global efficiency. The results showed that the global efficiency of the network in full-band EEG was significantly smaller in AD group for the values of T=0.06 and T=0.07, but there was no statistically significant difference in the clustering coefficients between the two groups for the values of T (0.05-0.07). However, the clustering coefficient and global efficiency were significantly lower in AD patients at alpha-band for the same threshold range than those of subjects in the control group. It suggests that there may be decreases of the brain connectivity strength in AD patients at alpha-band of the resting-state EEG. This study provides a support for quantifying functional brain state of AD from the brain network perspective.
All the collected original electroencephalograph (EEG) signals were the subjects to low-frequency and spike noise. According to this fact, we in this study performed denoising based on the combination of wavelet transform and independent component analysis (ICA). Then we used three characteristic parameters, complexity, approximate entropy and wavelet entropy values, to calculate the preprocessed EEG data. We then made a distinguishing judge on the EEG state by the state change rate of the characteristic parameters. Through the anesthesia and non-anesthesia EEG data processing results showed that each of the three state change rates could reach about 50.5%, 21.6%, 19.5%, respectively, in which the performance of wavelet entropy was the highest. All of them could be used as a foundation in the quantified research of depth of anesthesia based on EEG analysis.