Chinese Journal of Biomedical Engineering

Home

Editorial Board

Download

Quick Search

Office Online

Current Issue

Download Articles

Download

Links

2022 Vol. 41, No. 1 Published: 2022-02-20

Reviews Communications Regular Papers

Regular Papers

	1	Individualized Model Construction Method for Transcranial Magnetic Stimulation Navigation
		Zhang Yun, Ding Hui, Wang Guangzhi
		DOI: 10.3969/j.issn.0258-8021.2022.01.001
		Transcranial magnetic stimulation is a non-invasive stimulation technique widely used in the diagnosis and treatment of neurological and psychiatric diseases. Image-guided transcranial magnetic stimulation navigation can ensure the accurate positioning of the stimulation coil. However, some clinical contraindications make it difficult to obtain individual medical images of some patients, leading to the failure of image guidance. To solve this problem, this paper proposed to construct individualized images to replace individual images for navigation. First, we used positioning devices to acquire the patient's scalp contour points; and reconstructed the scalp surface based on Poisson reconstruction algorithm; then combined the automatic control points positioning method to locate control points; finally used the thin plate spline interpolation algorithm to construct individualized model. We tested the accuracy of the scalp, brain surface and brain targets of the individualized models. The results showed that the scalp accuracy reached (2.8±0.5) mm, and the brain surface accuracy reached (3.1±0.6) mm. The accuracy of the hand motor cortex target reached (9.2±4.1) mm, which met the needs of multi-target localization in the diagnosis of neurological diseases by transcranial magnetic stimulation navigation. The accuracy of the dorsolateral prefrontal lobe target reached (8.9±5.1) mm, which is better than the commonly used clinical alternative target (EEG electrode F3) whose accuracy is (16.1±5.0) mm. In conclusion, the accuracy can meet the need of repeated target localization in transcranial magnetic stimulation navigation for the treatment of mental diseases.
		2022 Vol. 41 (1): 1-9 [Abstract] ( 506 ) HTML (1 KB) PDF (6785 KB) ( 528 )

	10	Study on Topological Specificity of Resting-State Functional Brain Networks in EpileptogenicHemisphere of Temporal Lobe Epilepsy
		Cao Yingxin, Ge Manling, Chen Shenghua, Song Zibo, Xie Chong, Yang Zekun, Wang Lei, Zhang Qirui
		DOI: 10.3969/j.issn.0258-8021.2022.01.002
		Epilepsy is a typical neurological disease worldwide with abnormal neural discharges in the brain leading to dysfunction in the central cognitive functional networks. As an advanced technology today, the functional connectivity (fMRI-FC) derived from the resting-state functional magnetic resonance imaging (rfMRI) provides a scientific detection index for assessing the brain functions. Here, a fMRI-FC specificity model was proposed with reference to healthy individuals, based on multiple nodes indexes fusion in the whole brain functional networks in epilepsy, aiming to improve fMRI-FC detection to a high-order level. To validate the effectiveness, the model was employed to build the functional network topological metrics, and then applied to classify the epileptogenic hemisphere by a machine learning method. Firstly, the rfMRI data of a total of 20 mesial temporal lobe epilepsy patients, whose epileptogenic hemispheres were indicated by the positive hippocampal formation on the structure MRI (10 patients on each epileptogenic hemisphere) and a total of 139 healthy individuals were collected. Secondly, with FC as the edge, the brain functional networks were constructed. A total of 4 local nodes metrics were calculated for patients and healthy individuals. Thirdly, the fMRI-FC specificity model was constructed, with reference to the healthy individuals. The groups including 4 nodal indexes and 1 group of these indexes fusion were statistically employed to extract the sensitive brain areas to the epileptogenic hemisphere by ROC curve analysis, and the indexes of these areas were considered as the features to classify the epileptogenic hemisphere of the patients. The classification performance was analyzed by the leave-one-out method and random cross-validation. A fMRI-FC non-specific model was constructed by the multiple nodes indexes fusion of brain functional networks and was compared with the specific model built by us. The fMRI-FC specificity model of multiple nodes indexes fusion could classify the epileptogenic hemisphere effectively at an average classification accuracy of 95.0%±8.7%, that was validated by random cross-validation, and even 100% by leave-one-out method. The fMRI-FC specificity model of multiple nodes indexes fusion could effectively improve the localizing accuracy of epileptogenic hemisphere. Therefore, it might provide a new way for machine learning-aided assessing the epileptic brain by fMRI-FC.
		2022 Vol. 41 (1): 10-20 [Abstract] ( 391 ) HTML (1 KB) PDF (3441 KB) ( 433 )

	21	Early Detection of Depression Based on Textual Information in Social Media
		Zhang Mengna, Wang Junyan, Long Yang, Zhang Haofeng, Hu Yong
		DOI: 10.3969/j.issn.0258-8021.2022.01.003
		The traditional method of diagnosing depression is through face-to-face assessment and conversation. However, many patients with depression are reluctant to seek medical attention at an early stage, which makes their condition worse. In order to judge the situation of patients with depression in the early stage, a detection model using time series features of social media textual information and multi-instance learning was proposed in this work. Considering that depressive symptoms will not appear immediately, the use of time series samples will be very important. Therefore, the unsupervised LSTM was used to extract time series features, binary classification was implemented by training a classifier, and a multi-instance learning model was exploited to solve the problem of unbalanced samples. Naive Bayes classifiers, random forests, multivariate social network learning and multimodal depression dictionary learning were used as benchmark methods firstly, and then the multi-instance learning with unsupervised LSTM time series functions was employed to detect depression more accurately. On the basis of the MDDL dataset, 200 survey subjects totally 7946 tweets were selected, and the training-test ratio was set as 8:2. Experimental results were following: the accuracy, precision, recall and F1 score reached 75.0%, 76.0%, 73.0%, and 74.5%, respectively, which demonstrated that it was feasible to use machine learning for early depression detection through text data in social media. In addition, a large number of ablation studies also verified that the method using time series features could achieve better performance than the traditional benchmark methods.
		2022 Vol. 41 (1): 21-30 [Abstract] ( 521 ) HTML (1 KB) PDF (1628 KB) ( 1623 )

	31	Research on Intelligent Recognition Method of Common Arrhythmia Combining Traditionaland Deep Features of Single-Lead ECG
		Li Quanchi, Huang Xin, Luo Chengsi, Huang Huiquan, Rao Nini
		DOI: 10.3969/j.issn.0258-8021.2022.01.004
		Cardiac arrhythmia is one of the most common types of cardiovascular diseases. Long-term monitoring of few-leads ECGs based on portable devices is helpful to improve the detection rate of arrhythmias. But the large amounts of long-range ECG data generated impose a great burden on clinicians, which leads to missed detection and misjudgments. Therefore, this paper developed an automatic identification method for common arrhythmias by combining features of the traditional single-lead ECG with deep network features. The new method first extracted the traditional features in frequency domain, time domain, and morphology of common arrhythmias. Then a residual block deep convolutional neural network and a bidirectional long-short memory network were built to extract the deep network features. These three types of the features were fused in one deep network to classify heart rhythms including normal and arrhythmias. Finally, 6 877 sets of static and 8 528 sets of Holter data provided by the 2018 Chinese Physiological Signals Challenge and the 2017 Global Atrial Fibrillation Challenge were used to verify the method in this paper. With single-lead of static ECG signal, the method achieved an average F1 score of 0.855 for categorizing six arrhythmic rhythms and one normal rhythm, which is better than the existing relevant methods. As for single-lead dynamic ECG, the method achieved an average F1 score of 0.827 for categorizing AF, other arrhythmias, and normal rhythms, which is comparable to two methods tied for first in 2017 Global AF Challenge and superior to other related methods. Thus, this method has a good prospect of application in wearable remote monitoring and the auxiliary diagnosis of common arrhythmias.
		2022 Vol. 41 (1): 31-40 [Abstract] ( 357 ) HTML (1 KB) PDF (5324 KB) ( 493 )

	41	The Study of Lower Limb Motion Recognition Method Based on GA-RBF Neural Network andsEMG Signals
		Zhang Peng, Zhang Junxia, Liu Ruiheng, Ahmed Mohamed Moneeb Elsabbagh
		DOI: 10.3969/j.issn.0258-8021.2022.01.005
		To improve the accuracy of human surface electromyography (sEMG) signals for the recognition of lower limb movements, an RBF neural network classification model based on genetic algorithm (GA) optimization was proposed in this work. The sEMG of eight kinds of daily lower limb movements was collected, and the ‘sym6' wavelet function was selected for filtering preprocessing of sEMG. The principal component analysis (PCA) was used to reduce the dimension of time-frequency domain features, and the feature vectors were input into the RBF neural network optimized by GA for training and recognition. Experimental results showed that the average recognition rate of this method for the eight lower limb movements of the same subject was 94.00%±0.45 %, and the recognition rate for the lower limb movements of 15 different subjects reached 89.3 %, which was 11.8 % higher and 6 s shorter than that of the traditional BP neural network. The proposed method displayed a high recognition accuracy in the application of using sEMG signals to recognize human lower limb movements, providing a reference for the study of intention recognition of lower limb intelligent rehabilitation robot and of assistance in the rehabilitation of patients with lower limb disabilities.
		2022 Vol. 41 (1): 41-47 [Abstract] ( 328 ) HTML (1 KB) PDF (3902 KB) ( 525 )

	48	A Colorectal Segmentation Method Based on U-Net Improved with Identical Design
		Shen Zhiqiang, Lin Chaonan, Pan Lin, Nie Weiyu, Pei Yue, Huang Liqin, Zheng Shaohua
		DOI: 10.3969/j.issn.0258-8021.2022.01.006
		Colonoscopy is a widely used technique for colon screening and polyp lesions diagnosis. Nevertheless, manual screening using colonoscopy suffers from a miss rate around 25% of polyps. Deep learning-based computer-aided diagnosis (CAD) for polyp detection has potentials of reducing the human errors. Polyp detection depends on encoder-decoder network (U-Net) for polyp segmentation. However, U-Net has two limitations, one is that the semantic gap exists between the feature maps from the encoder and decoder; the other one is convolutional layers in the encoder-decoder processing units fail to extract multi-scale information. In this work, we proposed an identical network (I-Net) to tackle the problems in a consolidated manner. The I-Net introduced identical units (IU) both in skip connections and encoder-decoder sub-networks of U-Net to reduce the semantic gap. Meanwhile, motivated by the dense and residual connections, we designed a dense residual unit (DRU) to learn multi-scale information. Finally, DRI-Net was developed by initializing IU to DRU, which not only alleviated the semantic gap between the encoder and the decoder but also learned multi-scale features. We evaluated the proposed methods on the CVC-ClinicDB dataset containing 612 colonoscopy images through five-fold cross validation. Experimental results demonstrated that the DRI-Net achieved Dice coefficient of 90.06% and intersection over union (IoU) of 85.52%. Compared to the U-Net, DRI-Net improved the Dice coefficient of 8.50% and IoU of 11.03%. In addition, we studied the generalization of the proposed methods on International Skin Imaging Collaboration (ISIC) 2017 dataset including a training set of 2 000 dermoscopy images for model training and a test set of 600 images for model evaluation. The study indicated that the I-Net achieved Dice coefficient of 86.57% and IoU of 79.20%. Compared to the first-place solution on ISIC 2017 leaderboard, the DRI-Net improved Dice coefficient of 1.67% and IoU of 2.70%. In conclusion, the results demonstrated that DRI-Net effectively overcome the limitations of U-Net and improved the segmentation accuracy in the polyp segmentation task, and showed the great generalization capability on other modality data.
		2022 Vol. 41 (1): 48-56 [Abstract] ( 407 ) HTML (1 KB) PDF (7671 KB) ( 164 )

	57	Triple Attention Segmentation Network for Brain Tumor Images
		Han Yang, Song Jinmiao, Xue Anyi, Duan Xiaodong
		DOI: 10.3969/j.issn.0258-8021.2022.01.007
		Brain tumor segmentation is the basis of clinical routine and treatment of brain tumor diseases with computer-aided diagnosis. In this paper, we proposed a triple attention segmentation network based on brain tumor images, aiming to solve the problems of the current brain tumor MRI image segmentation network that has too many layers and is lack of connection between local and global feature information, which leads to the reduction of image segmentation accuracy. First, inspired by the residual network, we replaced the convolution module both in the encoding and decoding layer of the original segmentation network with a deep residual module to solve the problem of gradient disappearance caused by network deepening. Next, by introducing a triple attention module to combine local and global image features, the network was able to learn important image features better and improved the network's segmentation accuracy of brain tumor images. Finally, The improved network was evaluated by the Dice coefficient, and other brain tumor indicators were adopted on the BraTS brain tumor MRI image datasets released by the MICCAI competition includes 335 patient cases, among which the whole brain tumor score reached 85.20%, the brain tumor core score reached 87.10%, and the enhanced brain tumor area score reached 80.80%. Experimental results showed that the proposed segmentation network increased the segmentation performance of brain tumor MRI images without increasing the training time.
		2022 Vol. 41 (1): 57-63 [Abstract] ( 365 ) HTML (1 KB) PDF (3423 KB) ( 473 )

	64	Study of Ultrasonic Thyroid Nodules Detection Based on Cascade Rcnn
		Zhang Haowei, Li Zhanqi, Liu Ying, Li Miao
		DOI: 10.3969/j.issn.0258-8021.2022.01.008
		Thyroid ultrasound images have low contrast, unclear edges, high noise, and the surrounding tissues are complex and difficult to distinguish, making it extremely difficult for doctors to diagnose thyroid diseases. To overcome this problem, Cascade Rcnn target detection algorithm was used in this work, with ResNet50, Resnet101 and fusion compression incentive attention modules SE-ResNet50, SE-ReNet101 as the backbone network. There were 1 513 cases thyroid ultrasound images (including 832 cases benign nodules and 681 cases malignant nodules) obtained from a third-class hospital. Under the guidance of professional sonographers, the data were preprocessed into the standard coco format data set. The weights obtained from the pre-training of the large Imagenet database by transfer learning were migrated to this experimental model structure. Comparing with the experimental results of the four backbone networks, Cascade Rcnn algorithm with SE-ResNet101 as the backbone network achieved an accuracy of 92.4%, recall rate of 86.2%, specificity of 95.1%1, F1 value of 89.22%, and mAP value of 82.4%. The detection result of nodule localization and classification of benign and malignant was of clinical guiding significance for assisting doctors in the diagnosis of thyroid ultrasound images.
		2022 Vol. 41 (1): 64-72 [Abstract] ( 317 ) HTML (1 KB) PDF (6446 KB) ( 341 )

Reviews

	73	A Review of Research Progress of Hybrid Brain-Computer Interface
		Shi Wenqiang, Xiao Xiaolin, Liu Shuang, Xu Minpeng, He Feng, Ming Dong
		DOI: 10.3969/j.issn.0258-8021.2022.01.009
		Traditional brain-computerinterface (BCI) has many shortcomings in practical applications, such as a small instruction set, a small range of people, hard to achieve multi-dimensional control and asynchronous control. Hybrid Brain-Computer Interface (hBCI) can effectively solve these problems. In this paper, three common types of hBCI were reviewed, including hBCI based on EEG signals, hBCI based on EEG signals and other brain signals, and hBCI based on multiple physiological signals. In addition, this paper focused on the research status of hBCI systems and analyzed the stimulus paradigm, control strategy, classification performance, and practical application. The analysis results showed that compared with the traditional BCI system, the hBCI system has a much larger instruction set and higher accuracy. Moreover, due to the combination of other brain signals or physiological signals, hBCI is easier to realize multi-dimensional control and asynchronous control and has achieved rapid development in the utility performance of the system. Finally, this paper summarized different types of hBCI systems and proposed existing problems and future development prospects of hBCI.
		2022 Vol. 41 (1): 73-85 [Abstract] ( 533 ) HTML (1 KB) PDF (919 KB) ( 1189 )

	86	The Application Progress of Dynamic Causal Model in Brain Network Research
		Liang Sailan, Wang Duojin
		DOI: 10.3969/j.issn.0258-8021.2022.01.010
		The effective connectivity between different functional areas of the brain is one important research issue in brain science. It is of great significance to investigate the brain networks formed by effective connectivity between brain regions in different situations, which can help people to understand the comprehensive functional mechanism of the brain. This research also has advantages in the treatment of various brain-related diseases and the development of brain functions. Dynamic causal model (DCM) is an advantageous way to analyze effective connectivity in the brain network. In this paper, we reviewed the research on the dynamic causal model based on functional magnetic resonance imaging (fMRI), electroencephalogram (EEG), and functional near-infrared spectroscopy (fNIRS). The application of DCM in fMRI can be divided into stroke-related brain networks, cognitive neuroscience brain networks and mental disease related brain networks. The application of DCM in EEG mainly includes cognitive neuroscience and diseases related to schizophrenia, Alzheimer's disease, epilepsy, Parkinson's disease, etc., however, rare in fNIRS so far, is only involved with cognitive neuroscience. Finally, we compared the three technologies and discussed the prospectives.
		2022 Vol. 41 (1): 86-99 [Abstract] ( 723 ) HTML (1 KB) PDF (922 KB) ( 1063 )

	100	Review of Functional Magnetic Resonance Imagingin Diagnosis of Mild Cognitive Impairment
		An Xingwei, Zhou Yutao, Di Yang, Liu Shuang, Ming Dong
		DOI: 10.3969/j.issn.0258-8021.2022.01.011
		Nowadays Alzheimer's disease (AD) has severely influenced and limited personal daily life and even posed a grave threat to the life and health of patients. Mild cognitive impairment (MCI) is the prodromal stage of AD, and accurate diagnosis can help to interfere or reduce the conversion of patients to Alzheimer's disease. At present, functional magnetic resonance imaging (fMRI) technology have been widely used in the detection and diagnosis of MCI. This article introduced the research status of fMRI in MCI from the aspects of feature extraction, feature selection, data dimensionality reduction and classification recognition. First, the commonly used resolution indicators such as low-frequency amplitude, local consistency, and functional connection for feature extraction was introduced. Second, features selection and data dimension reduction methods were introduced, and the efficient machine learning and deep learning algorithms in classification and recognition were summarized. This paper also proposed the remained problems and made perspectives to the future research.
		2022 Vol. 41 (1): 100-107 [Abstract] ( 348 ) HTML (1 KB) PDF (1083 KB) ( 755 )

	108	Neuroimaging Research Progress of Negative Symptoms in Schizophrenia
		Gong Jinnan, Yang Zhuoru, Li Lu, Jiang Yuchao, Dong Debo, Shao Junming, Yao Dezhong, Luo Cheng
		DOI: 10.3969/j.issn.0258-8021.2022.01.012
		Patients with chronic schizophrenia have negative symptoms as the main clinical manifestations, which are also the main factor of mental disability, has a great effect on the life quality of patients. So far, the neural mechanism of negative symptoms is still unclear, making it hard to be controlled. This article reviewed the research progress in different aspects in this field, including the occurrence of schizophrenia associated to the damage of subcortical regions (the striatum and thalamus), and the abnormality of the subcortex-cortical connectivity associated to negative symptoms. This article also proposed the perspectives of the new method (reconstruction of topological brain connection) in the study of the mechanism of negative symptoms of schizophrenia, such as exploring the coupling of the dopaminergic neurotransmitter system of the striatum-thalamus-prefrontal loop, and the deconstruction of spatial distribution information. In addition, the application of deep learning can further decode the topological features of the critical brain connections relate to negative symptoms, which is expected to become a potentially approach to explore the mechanism of negative symptoms of schizophrenia.
		2022 Vol. 41 (1): 108-113 [Abstract] ( 271 ) HTML (1 KB) PDF (725 KB) ( 446 )

Communications

	114	*Phenotype Analysis of Pathogenic Single Nucleotide Polymorphism of Susceptibility Gene TNNC1* with Hypertrophic Cardiomyopathy Based on Bioinformatics**
		Li Jiangxi, Zhang Shimei, Wang Yuxing, Zhao Yue
		DOI: 10.3969/j.issn.0258-8021.2022.01.013

		2022 Vol. 41 (1): 114-118 [Abstract] ( 346 ) HTML (1 KB) PDF (2838 KB) ( 489 )

	119	Design, Manufacture and Evaluation of Multi-Roots Analogue Dental Implant
		Peng Wei, Xu Liangwei, Cheng Kangjie, You Jia, Yao Chunyan
		DOI: 10.3969/j.issn.0258-8021.2022.01.014

		2022 Vol. 41 (1): 119-123 [Abstract] ( 230 ) HTML (1 KB) PDF (6030 KB) ( 262 )

	124	Preparation and Functionalization of Magnetic Porous PLGA Microspheres
		Gu Jiayu, Zhang Chao, Shi Lele, Xia Yan
		DOI: 10.3969/j.issn.0258-8021.2022.01.015

		2022 Vol. 41 (1): 124-128 [Abstract] ( 314 ) HTML (1 KB) PDF (4251 KB) ( 487 )

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