The brain-computer interface (BCI) based on motor imagery electroencephalography (MI-EEG) enables direct information interaction between the human brain and external devices. In this paper, a multi-scale EEG feature extraction convolutional neural network model based on time series data enhancement is proposed for decoding MI-EEG signals. First, an EEG signals augmentation method was proposed that could increase the information content of training samples without changing the length of the time series, while retaining its original features completely. Then, multiple holistic and detailed features of the EEG data were adaptively extracted by multi-scale convolution module, and the features were fused and filtered by parallel residual module and channel attention. Finally, classification results were output by a fully connected network. The application experimental results on the BCI Competition IV 2a and 2b datasets showed that the proposed model achieved an average classification accuracy of 91.87% and 87.85% for the motor imagery task, respectively, which had high accuracy and strong robustness compared with existing baseline models. The proposed model does not require complex signals pre-processing operations and has the advantage of multi-scale feature extraction, which has high practical application value.
Ballistocardiogram (BCG) and electrocardiogram (ECG) can realize the detection of cardiac function from mechanical and electrical dimensions respectively. By extracting the corresponding characteristic parameters of the two signals and carrying out joint analysis, an important cardiac physiological index such as cardiac contractility, can be reflected. To overcome the shortcomings of complication and heaviness of the existing acquisition equipment, a wearable BCG-ECG signal acquisition system is designed in this paper, which realizes BCG signal acquisition based on accelerometer and ECG signal acquisition based on conductive rubber electrodes. The signals of 6 healthy persons were collected, and BCG signals collected by piezoelectric films were used as reference signals. The waveform characteristics of signals were compared, and the difference of cardiac cycle acquisition was analyzed. The waveform characteristics of the two signals acquired by the device were consistent with the standard signals, and there was no significant difference in the acquisition of the cardiac cycle between the proposed method and the traditional method. The results show that the system can accurately collect human BCG signals and ECG signals. The system provides a basis for subsequent research on BCG signal formation mechanism and health applications.
Motor imagery electroencephalogram (EEG) signals are non-stationary time series with a low signal-to-noise ratio. Therefore, the single-channel EEG analysis method is difficult to effectively describe the interaction characteristics between multi-channel signals. This paper proposed a deep learning network model based on the multi-channel attention mechanism. First, we performed time-frequency sparse decomposition on the pre-processed data, which enhanced the difference of time-frequency characteristics of EEG signals. Then we used the attention module to map the data in time and space so that the model could make full use of the data characteristics of different channels of EEG signals. Finally, the improved time-convolution network (TCN) was used for feature fusion and classification. The BCI competition IV-2a data set was used to verify the proposed algorithm. The experimental results showed that the proposed algorithm could effectively improve the classification accuracy of motor imagination EEG signals, which achieved an average accuracy of 83.03% for 9 subjects. Compared with the existing methods, the classification accuracy of EEG signals was improved. With the enhanced difference features between different motor imagery EEG data, the proposed method is important for the study of improving classifier performance.
ObjectiveTo investigate the correlation and clinical significance of 8-hydroxydeoxyguanosine (8-OHdG) and endothelin-1 (ET-1) levels with cognitive dysfunction in patients with chronic obstructive pulmonary disease (COPD), and provide new idea for the prevention and treatment for cognitive dysfunction in COPD patients.MethodsA total of 103 COPD patients, according to the Montreal cognitive assessment scale standard for evaluation, were divided into a cognitive dysfunction group and a cognitive normal group. Serum 8-OHdG and ET-1 levels were compared between the two groups and their correlations with cognitive function were analyzed with the receiver operating characteristic (ROC) curve.ResultsThe levels of serum 8-OHdG and ET-1 in the COPD patients with cognitive impairment were significantly higher than those in the cognitive normal group ［8-OHdG: (13.91±9.04) ng/ml vs. (7.28±3.00) ng/ml; ET-1: (95.64±57.66)pg/ml vs. (69.20±7.89)pg/ml］ (both P<0.05). The levels of 8-OHdG (OR=22.94, 95%CI 7.06-74.53) and ET-1 (OR=19.76, 95%CI 6.59-59.31) were associated with cognitive impairment in the COPD patients. The areas under ROC curve of serum 8-OHdG and ET-1 levels to predict cognitive dysfunction in the COPD patients were 0.786 (95%CI 0.691-0.881) and 0.790(95%CI 0.695-0.885).ConclusionsThe serum levels of 8-OHdG and ET-1 are associated with cognitive impairment in COPD patients. The levels of 8-OHdG and ET-1 in serum can predict cognitive impairment with high specificity.
ObjectiveTo investigate the expression of C/EBP homologous protein (CHOP) in lung tissue of chronic intermittent hypoxia rats, and explore the intervention effect of edaravone and its possible mechanism.MethodsA total of 120 adult male Wistar rats were randomly divided into three groups: a normal control group (UC group), a chronic intermittent hypoxia group (CIH group), an edaravone intervention group (NE group), and a normal saline group (NS group). The above four groups were also randomly divided into five time subgroups of 3 days, 7 days, 14 days, 21 days and 28 days, respectively, with 6 rats in each time subgroup. The histopathological changes of lung tissue were observed by hematoxylin-eosin (HE) staining and the expression of CHOP in lung tissue was detected by immunohistochemical method.ResultsHE staining results showed that there was no obvious pathological change in UC group. The epithelial cells of lung tissue in CIH group showed edema, hyperemia, widening of alveolar septum and inflammatory cell infiltration. The pathological injury was more serious with the prolongation of intermittent hypoxia time. There were also pathological changes in NE group, but the degree of lung tissue injury was significantly lower than that in CIH group. The results of immunohistochemistry showed that the expression of CHOP in CIH group was significantly higher than that in UC group. The expression of CHOP in NE group was higher than that in UC group, but it was still significantly lower than that in CIH group.ConclusionsThe expression of CHOP protein in lung tissue of chronic intermittent hypoxic rats is enhanced and the high expression of CHOP protein plays a certain role in the lung injury of chronic intermittent hypoxia rats complicated with lung injury. Edaravone may protect lung tissue from chronic intermittent hypoxia by inhibiting the expression of CHOP.
Objective To observe the relationship of serum levels of homocysteine (HCY) and chemokine C-C motifligand 2 (CCL2) with cognitive impairment in COPD patients with different degrees of emphysema. Methods Sixty-twoCOPD patients identified according to emphysema phenotype classification and admitted from January 2016 to March 2017 were recruited in the study. There were 37 cases in emphysema 1-2 grade and 25 cases in emphysema 3-4 grade. Simultaneous 30 healthy subjects undergoing physical examination were recruited as control. Montreal cognitive assessment (MoCA) scale investigation and serum HCY and CCL2 test were completed. Relationship analysis was conducted on serum HCY, CCL2 levels with cognitive impairment in the COPD patients with different degrees of emphysema. Results Compared with the 1-2 grade subgroup, the PaO2 was lower, PaCO2 was higher, the plasma HCY and CCL2 levels increased in the 3-4 grade subgroup with significant differences (all P<0.05). MoCA total score and subscores were relatively low in the COPD group with emphysema than the control group (except visuospatial ability scores in the 1-2 grade subgroup). MoCA scores were statistically lower in the 3-4 grade subgroup than those in the 1-2 grade subgroup (allP<0.05). Correlation analysis showed that HCY and CLL2 levels were negatively correlated with MoCA scores and subscores (P<0.01), and HCY and CLL2 were positively correlated (bothP<0.01). The area under the receiver operating characteristic curve of HCY and CLL2 for evaluating cognitive impairment was 0.79 and 0.97, respectively. Conclusion In patients with different degrees of emphysema phenotype, serum HCY and CCL2 levels are increased in different degree, and the degree of emphysema is closely related with cognitive dysfunction.