Chinese Journal of Biomedical Engineering

Home

Editorial Board

Download

Quick Search

Office Online

Current Issue

Download Articles

Download

Links

2023 Vol. 42, No. 4 Published: 2023-08-20

Reviews Regular Papers

Regular Papers

	385	Cortico-Muscular Coupling Analysis Based on Cumulative Spike Sequence of Motor Unit
		Su Jiahao, She Qingshan, Zhang Jianhai, Ma Yuliang, Fan Yingle
		DOI: 10.3969/j.issn.0258-8021.2023.04.001
		Cortico-muscular coupling can reflect the connection between cerebral cortex and muscle in sensorimotor. This paper proposed a new cortico-muscular coupling analysis method, that is utilizing EEG and cumulative spike train (CST) obtained after the decomposition of the motor unit to linearly transmit the neural drive for coherence analysis, quantitatively assessing cortico-muscular coupling and common synaptic input of neurons under different frequency band at different contraction force levels during upper limb grasping. The synchronous EEG and sEMG data of flexor digitorum superficialis (FDS) and flexor carpi ulnaris (FCU) were measured and analyzed in 10 healthy subjects. Results showed that both frequency band (F (4, 8)=337.2, P<0.01) and contractile force level (F (2,8)=12.15, P<0.01) had significant effects on the intermuscular coupling during upper limb grasping exercise, especially in β and α band. At 30% MVC, the mean coherence of β frequency band was 0.23 ± 0.10, and that of α frequency band was 0.47 ± 0.02. The synaptic input controls the level of contraction force. Cortico-muscular coupling was relatively low. The highest coupling strength was in β band with a coherence value of 0.12 ± 0.01 at 30% MVC. The CST brain muscle coupling analysis can reflect the coupling characteristics and common synaptic inputs of various frequency bands and contraction levels between brain muscles, providing a new method for brain muscle coupling analysis.
		2023 Vol. 42 (4): 385-393 [Abstract] ( 322 ) HTML (1 KB) PDF (3600 KB) ( 203 )

	394	Effects of Transcranial Direct Current Stimulation on Features of Human Balance Brain Network
		Cai Hongwei, Luo Zhizeng, Shi Hongfei
		DOI: 10.3969/j.issn.0258-8021.2023.04.002
		Transcranial electrical stimulation is a neuromodulation technology with wide application potential, which still remains on the external manifestation in the research of human balance, and mechanisms of influence on the neuromodulation of human balance have not been clear yet. Brain functional network is an effective means to understand brain function and regulation mechanism. In this work, a method to study the features of human balanced brain network by transcranial electrical stimulation was proposed. There were 24 subjects recruited to apply sham/ anodal/cathodal transcranial direct current stimulation (tDCS) in the primary motor cortex, and the planar center of pressure (COP) and EEG data were collected while standing with eyes closed, and the features of COP and the features of network including average degree, clustering coefficient, path length and global efficiency were compared and analyzed. Experimental results showed that anodal tDCS reduced human shaking frequency. The average degree, clustering coefficient, path length and global efficiency after sham tDCS and anodal tDCS were 6.11±1.21 and 0.51±0.07, 1.85±0.16 and 0.64±0.05, 7.46±1.05, 0.61±0.06, 1.66±0.14, 0.69±0.04, respectively. There were significant differences in all groups (P<0.05). The shaking frequency was increased by cathodal tDCS, and the average degree, clustering coefficient, path length and global efficiency were 5.51±1.33, 0.43±0.07, 1.95±0.16 and 0.62±0.04, respectively, which were significantly different from those after sham tDCS (P<0.05). These results showed that anodal tDCS effectively activated the activity of motor cortex, enhanced the inter-interval connection of the whole brain network, and improved the balance ability of the human; while cathodal tDCS inhibited the activity of the motor cortex, and the balance ability of the human was weakened.
		2023 Vol. 42 (4): 394-402 [Abstract] ( 147 ) HTML (1 KB) PDF (2833 KB) ( 180 )

	403	Attention Enhancement and its Neuroplasticity Based on Long-Term Video Game Training
		Bai Binnan, Tian Xiaoyan, Cui Ruifang, Hao Xinyang, Lin Liqi, Gong Diankun, Yu Zhenxia, Gao Dongrui
		DOI: 10.3969/j.issn.0258-8021.2023.04.003
		It has been shown that action-based video game training can improve multiple cognitive abilities. However, current existing video game training games are of a single type and short duration. To fill this gap, a study was conducted to improve attention with prolonged training of multiple types of video games (GSGO, LOL and SGS). In this study, 176 healthy undergraduate students were randomly divided into 3 training groups to receive game training for a period of 5 months. Three behavioral tasks were completed every month during the training period to assess attentional capacity, and EEG data were collected once before and after the training period to calculate brain power spectrum energy to assess changes in brain functional status after long-term video game training. The results revealed that after 5 months of game training, the θ power spectrum energy significantly increased in each training group and differed significantly in the parietal lobe (F=3.13, P<0.05). α power spectrum energy in the LOL and SGS groups showed a decreasing trend in all cortices, but the CSGO group showed a decrease in frontal (t=2.43, P=0.02), parietal (t=2.28, P=0.03) and central regions (t=2.48, P=0.02) showed significant α1 synchronization. There were significant between-group differences in R3 power spectrum energy in the frontal and parietal lobes across training groups (F=3.69, P=0.03), with both the LOL and SGS groups showing a significant increase in R3 power spectrum energy compared to pre-training (P<0.05), indicating that long-term electronic training can induce an increase in the attention levels and that the effect differed after training for different types of games. The results of the behavioral task assessment supported the same findings, with all training groups showing significantly higher performance in distracted spatial attention ability after month 1 of training (P<0.05), with the CSGO group improving by 11.12% and the SGS and LOL groups improving by 9.68% and 15.66%, respectively. Multiple comparisons of LSD showed that the CSGO group performed significantly better during training than the SGS group (P=0.01) and the LOL group had significantly better focused spatial attention than the CSGO group after the first 3 months of training (P=0.02). The study demonstrated that video games causally enhanced attentional plasticity and that action video games had better attentional capacity enhancement effects compared to casual game training. The results of the study are informative for assessing the effects of video game training and intervention.
		2023 Vol. 42 (4): 403-410 [Abstract] ( 193 ) HTML (1 KB) PDF (2175 KB) ( 300 )

	411	Effects of Duration of Transcranial Direct Current Stimulation on Brain Motor Cortex Activity
		Huang Fuxin, Sun Yao, Li Jingqi, Luo Zhizeng
		DOI: 10.3969/j.issn.0258-8021.2023.04.004
		Transcranial direct current stimulation (tDCS) is an emerging non-invasive brain stimulation technique, which can enhance cortical excitability and thus promote motor rehabilitation. However, the criteria for setting the stimulation duration parameters of this technique in motor function rehabilitation training have not been clearly defined. In this paper, the effect of anodal tDCS duration on the activity of cortical motor areas was investigated by acquiring EEG signals with a constant stimulation DC intensity of 1.5 mA. In the experiment, 11 healthy subjects were selected to receive three different durations (10, 15, 20 min) of anodal stimulation in C3 area based on the empirical values of previous experiments. In addition, a set of 25 min same-mode stimulation experiments was added. The EEG signals were collected before and after the stimulation, and two characteristic parameters, event-related desynchronization of C3 area mu and beta rhythms during motor imagery and power spectral density in resting state, were measured to compare the effects of the four different duration of anodal stimulation on the activity of motor areas in cerebral cortex. Results showed that the change in power spectral density (μV²/Hz) of mu and beta rhythm before and after different length of stimulation was 10 min: 2.835±0.841, 15 min: 3.975±0.978, 20 min: 7.022±1.562, 25 min: 1.413±1.329and 10 min: 0.890±0.421, 15 min: 1.645±0.630, 20 min: 3.122±0.710, 25 min: 0.321±0.259. The change in event-related desynchronization (%) of mu and beta was 10 min:10.06±2.81, 15 min: 14.11±2.87, 20 min: 22.12±4.67, 25 min: 3.77±3.03 and 10 min: 6.72±3.19, 15 min: 11.78±5.02, 20 min: 17.81±3.16, 25 min: 2.54±2.30. It was shown that within the 10~20 min experiments, the longer the stimulation time, the stronger the brain motor cortex activity was in the range of the 3 empirical values, and there was a significant difference (P<0.05), but continued prolongation of the stimulation duration instead led to a decrease in the brain motor cortex activity. These results are expected to provide a meaningful reference for the stimulation duration setting of motor function rehabilitation training based on tDCS technique.
		2023 Vol. 42 (4): 411-419 [Abstract] ( 140 ) HTML (1 KB) PDF (3263 KB) ( 130 )

	420	The Test-Retest Assessment of Apparent Fiber Density Measurements Using Diffusion Weighted Imaging in English
		Luo Yichao, Yan Jingguo, Chen Yuanyuan, Fan Qiuyun
		DOI: 10.3969/j.issn.0258-8021.2023.04.005
		Diffusion weighted imaging has been widely used in quantitative study of human brain white matter. Apparent fiber density (AFD) is a quantitative metric based on the fiber orientation distribution (FOD), which can reflect the fiber density of white matter and has been applied in the research of healthy development and neurodegenerative diseases. However, the value of AFD measurement strongly depends on experimental set-up, and the test-retest repeatability of AFD under different experimental conditions has not been systematically assessed. In this paper, the test-retest repeatability of AFD was examined at different diffusion weighted values (b-values) using the scan-rescan data of 42 subjects from two separate datasets. One dataset (n=35) is from Human Connectome Project (HCP), the other (n=7) is form Connectome Diffusion Microstructure Dataset (CDMD). The AFD value of fixel level was estimated using fixel-based analysis (FBA), and tract ROIs were obtained using TractSeg to estimate the fiber bundle averaged AFD. Results showed that at the fixel level, as the b-value increased, the Pearson’s correlation coefficient (r) and intraclass correlation coefficient (ICC) increased from 0.796 9 and 0.898 5 to 0.882 8 and 0.941 4 respectively in the test-retest experiment with HCP, and the absolute deviation decreased from 0.124 1 to 0.107 3. Similar trend was also found in the dataset from CDMD. At the tract-average level, the test-retest repeatability can be achieved at all b-values in both datasets. In summary, this paper examined the repeatability of AFD estimation for b-values in the range of 1 000~5 000 s·mm^-2, and our results might provide useful insights into the experimental design in future neurobiological investigations using AFD.
		2023 Vol. 42 (4): 420-430 [Abstract] ( 121 ) HTML (1 KB) PDF (6094 KB) ( 82 )

	431	Application of LMD-UNET Network in Multi-Modal MRI Images Segmentation of Brain Tumors
		Xia Jingming, Tan Ling, Liang Ying
		DOI: 10.3969/j.issn.0258-8021.2023.04.006
		There is a semantic gap among the feature maps corresponding to the codec in the UNet network, and its dual roll integration layer cannot learn multi-scale information, resulting in the loss of some feature information, which affects the MRI image segmentation effect. To solve this problem, this paper proposed a new image segmentation network local residual fusion multi-scale dual branch network LMD-UNet. In the coding process, the network used local feature residuals to fuse dense blocks and multi-scale convolution modules to expand the receptive field of images and optimize the propagation of underlying visual features; and in the decoding process, the network used double branch convolution to generate new high-level semantic features to reconstruct the information lost in the coding path. For segmentation experiments, 335 cases of the public brain tumor dataset BraTs were used, and the segmentation results were compared with U-Net that is currently a mainstream segmentation network. Experimental results showed that the four objective evaluation indexes of LMD-UNet model, precision, dice, 95% HD and recall reached0.933, 0.921, 0.702 and 0.966 respectively. Compared to U-Net, the corresponding indicators increased by 6.3%, 5.7%, 1.8%, and 6.1%, respectively, which indicated that LMD-UNet achieved more precise segmentation of brain tumor images. Meanwhile, the proposed method also showed a good performance in the edge contour segmentation for the detail part, which prospectively provided guarantee for the diagnosis of brain tumor and the surgery.
		2023 Vol. 42 (4): 431-441 [Abstract] ( 143 ) HTML (1 KB) PDF (4195 KB) ( 192 )

	442	Multi-Channel Sparse Graph Transformer Network for Early Identification of Alzheimer′s Disease
		Qiu Yali, Zhu Yun, Yu Shuangzhi, Song Xuegang, Wang Tianfu, Lei Baiying
		DOI: 10.3969/j.issn.0258-8021.2023.04.007
		Currently, there is no effective treatment for Alzheimer's disease (AD). Clinical intervention via early diagnosis can delay the AD progression and improve its prognosis. However, the existing methods only consider the neuroimaging features learned from group relationships, not the individual characteristics of the subjects. In this work, we designed a novel multi-modal multi-channel sparse graph transformer network (MSGTN). Our proposed network model included two parts, they are multi-modal data optimization and multi-modal feature learning. Firstly, we acquired the image information (e.g., diffusion tensor imaging (DTI) and functional magnetic resonance imaging (fMRI)) and non-image information (e.g., age and sex) of each subject. Secondly, we utilized locally weighted clustering coefficients (LWCC) to fuse functional and structural information. After that, the fused multi-modal image features were combined with the gender and age information of the subjects to construct a sparse graph. Finally, we input the sparse graph into the MSGTN network for early AD identification. We obtained a total of 170 subjects from the public database ADNI (Alzheimer's disease neuroimaging initiative), including 38 LMCI, 44 EMCI, 44 SMC, and 44 normal controls (NC). Our method achieved classification accuracy of 87.02%, 87.40%, 91.49%, 88.93%, 86.74% and 92.12%, respectively. The experimental results have proved that our proposed model not only can analyze NC versus three different early AD disease states, but also achieved superior classification performance in three different early AD disease states.
		2023 Vol. 42 (4): 442-452 [Abstract] ( 137 ) HTML (1 KB) PDF (4076 KB) ( 70 )

	453	Drug-Disease Association Prediction Based on Multi-Feature Fusion
		Kang Hongyu, Li Qin, Li Jiao, Gu Yaowen, Hou Li
		DOI: 10.3969/j.issn.0258-8021.2023.04.008
		We constructed a drug-disease association prediction model on the basis of drug multi feature fusion, which can provide theoretical foundation for drug knowledge discovery. Three similarities fused into drug comprehensive similarity by drug chemical structure, drug-side effect and drug-target multi-features. Disease similarity was calculated based on MeSH tree number. Next, GCN method was used to extract feature information of drug-disease graph data. Finally, MFFGCN was constructed for drug-disease association prediction. The association of drug diseases was predicted on the same data set, with the help of multiple evaluation indicators such as AUC, AUPR, accuracy, sensitivity, recall and F1, MFFGCN has better performance than the single feature association prediction method and 4 existing representative algorithms. The AUC index is 0.866 2, which is 2.48% higher than the average predicted AUC index of single feature and 1.67% higher than the baseline method. The AUPR index is 0.3412, which is 1.67% higher than the average predicted AUC index of single feature and 27.49% higher than the baseline method. MFFGCN has achieved good performance in the prediction of unknown drug disease association. This methods can find new indications of drugs, and also provide methodological reference and theoretical basis for drug relocation.
		2023 Vol. 42 (4): 453-460 [Abstract] ( 139 ) HTML (1 KB) PDF (1905 KB) ( 176 )

Reviews

	461	Research Progress of High Frequency Electrocardiography
		Xing Yantao, Li Jiayi, Xiao Zhijun, Li Jianqing, Liu Chengyu
		DOI: 10.3969/j.issn.0258-8021.2023.04.009
		The invention of the electrocardiogram provides rich information for the diagnosis, prevention, and treatment of cardiovascular diseases, but the information behind the electrocardiogram is far more than that. The new generation of high-frequency ECG technology can mine deeper human physiological information and further make up for the shortcomings of conventional ECG in the diagnosis of myocardial ischemia and the evaluation of autonomic nerves system status. Starting from the principle and development history of high-frequency ECG technology, this paper summarized the key technologies of high-frequency ECG technology in four aspects including signal sensing, acquisition system, signal denoising, and feature engineering. Based on the development of these technologies, this paper further summarized the achievements and applications of high-frequency ECG technology in the clinical practice of myocardial ischemia and autonomic nerves system status assessment, and analyzed and discussed the technical limitations of high-frequency ECG in practical application and the future development direction.
		2023 Vol. 42 (4): 461-474 [Abstract] ( 139 ) HTML (1 KB) PDF (4069 KB) ( 140 )

	475	Application of Neuroimaging-Based Deep Learning Model Interpretability Methods in Alzheimer′s Disease Recognition
		Wang Jiarong, Ke Ming, Dong Zhanguo, Wang Lubin, Li Liang
		DOI: 10.3969/j.issn.0258-8021.2023.04.010
		Alzheimer′s disease (AD) is a neurodegenerative disease. So far the pathogenesis has not been completely understood, and it can cause severe life and economic burden to the family and society after the onset. Timely prevention and intervention can delay the occurrence and development of AD. There is no complete cure, so early screening for AD is of great significance. This review paper firstly summarized the AD prediction model based on deep learning technology and compared and analyzed the model structure, data scale, and local and global brain regions. Among them, the 3D convolutional neural network model showed the best performance, and the continuous expansion of the data scale was helpful to improve the model performance. This paper also summarized the interpretability methods of the current neuroimage model, analyzed the advantages and disadvantages of the interpretability methods based on sensitivity analysis and backpropagation in the application of AD diagnosis, and demonstrated the interpretability method represented by backpropagation was more suitable for AD research. Finally, according to the research status, next development direction was proposed, which was suggested to realize semantic medical image analysis and generate understandable diagnostic reports.
		2023 Vol. 42 (4): 475-485 [Abstract] ( 149 ) HTML (1 KB) PDF (1815 KB) ( 318 )

	486	Progress in the Development of Man-Machine-Environment Integrated Intelligent Prosthetic Knee
		Wang Xiaoming, Li Linrong, Chen Changlong, Sun Jie, Zhang Zhewen, Meng Qiaoling, Yu Hongliu
		DOI: 10.3969/j.issn.0258-8021.2023.04.011
		The intelligent prosthetic knee, as the most important component in the lower limb prosthesis system, is a sophisticated system with the high integration of man-machine-environment. The solution of its key programs is a crucial step towards natural and compliant gait of the amputee wearing a prosthetic knee in practical applications. In this article, we reviewed the research progress of the intelligent prosthetic knee from the perspectives of bionic design, intelligent perception, and intelligent control. Bionic design focuses on how to make the prosthetic knee fit the driving / damping compensation mechanism of human joint through bionic structure and actuator design. Intelligent perception focuses on how to establish a man-machine-environment interaction channel to realize the hybrid decision-making in locomotion intention recognition and man-machine collaborative tasks. Intelligent control focuses on how to adjust the actuating strategy of the prosthetic knee in a dynamic environment to improving the approximation with normal gait characteristics. Finally, we discussed future directions and challenges in the research of the intelligent prosthetic knee, including the hybrid active-passive compensatory actuating, volitional control and man-machine-environment double closed-loop interaction.
		2023 Vol. 42 (4): 486-501 [Abstract] ( 121 ) HTML (1 KB) PDF (3530 KB) ( 133 )

	502	Advances in the Research on Induced Differentiation of Pluripotent Stem Cells into HematopoieticStem Progenitor Cells
		Li Yanwei, Shan Wei, Liu Li, Huang Qiong, Fang Sanhua
		DOI: 10.3969/j.issn.0258-8021.2023.04.012
		Hematopoietic stem cells (HSC) transplantation is an effective means for the treatment of various hematological diseases. At present, allograft and autologous HSC transplantation are widely used in clinical practice, but there are still many problems, such as graft-versus-host disease caused by allogeneic transplantation, the limited number of HSC for autologous HSC transplantation. Pluripotent stem cells (PSC) including embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs) have the ability of self-renewal and multidirectional differentiation in vitro, which theoretically continuously produce HSC. However, there are still key scientific problems to be solved in the study of PSC hematopoietic differentiation, such as low efficiency in hematopoietic differentiation and difficulties to obtain PSC derived HSC that has long-term hematopoiesis potential in vivo. The reasons are that the internal regulation mechanism and external environment for PSC hematopoietic differentiation has not clear yet. This article reviewed the role of internal regulatory elements including transcription factors and signaling pathways, and external microenvironmental factors including stromal cells, cytokines and novel biomaterials in regulating PSC hematopoietic differentiation during in vitro and in vivo embryonic hematopoietic development, providing a theoretical basis for the study of PSC hematopoietic differentiation in vitro and clinical transformation of PSC-derived HSC.
		2023 Vol. 42 (4): 502-512 [Abstract] ( 143 ) HTML (1 KB) PDF (2309 KB) ( 289 )

Copyright © Editorial Board of Chinese Journal of Biomedical Engineering
Supported by:Beijing Magtech