Chinese Journal of Biomedical Engineering

Home

Editorial Board

Download

Quick Search

Office Online

Current Issue

Download Articles

Download

Links

2019 Vol. 38, No. 5 Published: 2019-10-20

Reviews Communications Regular Papers

Regular Papers

	513	Automatic Segmentation of Hepatic Vein and Portal Vein Based on W-Net
		Sun Jinfeng, Ding Hui, Wang Guangzhi
		DOI: 10.3969/j.issn.0258-8021.2019.05.001
		Automatic segmentation of hepatic vein and portal vein based on preoperative CT images has important clinical value for liver segment. However, in the venous CT image of the liver, little difference is in the density of the hepatic vein and portal vein, and the vascular structure is also complicated, so there has been a challenge with automatic extraction of the hepatic vein and portal vein in 3D. To solve this problem, this paper proposed a network architecture named W-Net based on convolutional neural network (CNN). The network architecture made a full use of the difference in the 3D structure for the hepatic vein and portal vein. We also set a loss function for the extraction of all vessels and portal vein. Then the model would optimize the weighted sum of the two loss functions to automatically learn the features of all vessels and portal vein to achieve the best extraction result. Hepatic vein was obtained by subtracting the two result. In this paper, we used 10 sets of venous abdominal CT images from the public data set 3Dircadb01 for network model construction, and 10 groups for testing. We mainly used the Dice coefficient as the evaluation standard. Finally, in the liver area, the Dice coefficient and accuracy of blood vessels reached 0.715 and 0.970, the Dicecoefficient and accuracy of hepatic vein reached 0.597 and 0.984, and the Dice and accuracy of portal vein reached 0.608 and 0.970. We also tested in 10 sets of clinical data, and the method could effectively separate the hepatic vein and portal vein. The experimental results showed that the proposed method possessed better feature extraction ability and generalization ability, and had good performance in public data and clinical data.
		2019 Vol. 38 (5): 513-521 [Abstract] ( 800 ) HTML (1 KB) PDF (7569 KB) ( 407 )

	522	Small Intestinal Polyp Detection in Wireless Capsule Endoscopy Images
		Fan Shanhui, Liu Shichen, Cao E, Fan Yihong, Wei Kaihua, Li Lihua
		DOI: 10.3969/j.issn.0258-8021.2019.05.002
		Polyp is one of the most common small intestinal diseases. Wireless capsule endoscopy (WCE) is one of the routine clinical methods used for intestinal disease diagnosis. WCE can produce a mass of images during one examination, which may only contain a few abnormal images, therefore it is time consuming for doctors to review those images, which easily causes false detection and/or missed detection. Therefore, the development of an automatic polyp detection method is greatly valuable to provide support for doctors with better accuracy and efficiency. This study proposed a novel framework combining deep learning, transfer learning and data augmentation methods for polyp detection. The dataset used for model training and evaluation contained6 920 normal images and 6 864 polyp images, which was augmented from an original dataset containing 4 300 normal images and 429 polyp images. Specifically, three convolutional neural networks varied from depth to depth (AlexNet, VGGNet and GoogLeNet) were trained from scratch. The results showed that the GoogLeNet achieved the best performance with a sensitivity of 97.18%, specificity of 98.78% and accuracy of 97.99%. However, the training of deeper networks required more time and better computer, so we performed transfer learning strategy by fine-tuning a pretrained AlexNet. This model achieved a high accuracy of 97.74%, sensitivity of 96.57%, specificity of 98.89% and the area under the receiver operating characteristic curve (AUC) of 0.996. The proposed method provided an effective way for precise automatic intestinal polyp detection with limited training data, lower time cost and computer configuration, having potentials to help doctors efficiently detect intestinal polyp with WCE images.
		2019 Vol. 38 (5): 522-532 [Abstract] ( 644 ) HTML (1 KB) PDF (11804 KB) ( 154 )

	533	Pediatric Left Cardiac Echocardiography Segmentation via BiSeNet
		Hu Yujin, Lei Baiying, Guo Libao, Mao Muyi, Jin Zelong, Chen Siping, Xia Bei, Wang Tianfu
		DOI: 10.3969/j.issn.0258-8021.2019.05.003
		Accurate segmentation of pediatric echocardiography is an essential step for the later biomedical measurement and diagnose. Currently, it relies on sonographer′s manual segmentation, which is time consuming and redundant, and therefore may lead to mistakes. Deep learning methods have achieved remarkable results in the field of computer vision. Therefore, we proposed to extract features from pediatric echocardiography images via deep convolutional neural networks and segment key anatomical structures of the heart. Specifically, we used BiSeNet consisting of two components, spatial path and context path, to extract low and high level features, respectively, and then fused them via a feature fusion module to get the most important features, for accurate segmentation. We conducted experiments on a dataset consisting of 87 echocardiography videos (2216 images) collected from Shenzhen Children Hospital, and compared our prediction with sonographers′ ground truth. Results showed that BiSeNet was able to capture the structure feature of echocardiography images, and achieved 0.914 and 0.887 in term of Dice index in left ventricle and left atrium segmentation task, respectively. The proposed method could help with accurate pediatric echocardiography segmentation, and released sonographers from redundant work.
		2019 Vol. 38 (5): 533-539 [Abstract] ( 505 ) HTML (1 KB) PDF (16635 KB) ( 326 )

	540	Change Detection Based on Sparse Representation for Retina Fundus Image Pair
		Fu Yinghua, Li Jiang, Pan Dongyan, Wang Guozheng, Fu Dongxiang
		DOI: 10.3969/j.issn.0258-8021.2019.05.004
		Change detection with a pair of retinal fundus images was focused on comparing two images captured at different stages. Illumination variations between the image pair, along with the intensity similarity of anatomic structures and lesions make the pixel-by-pixel methods based on subtraction operation or ratio operation hard to obtain clear changing areas. In this paper, a new change detection method based on sparse representation classification (SRC) was proposed, aiming to reduce the illumination variations between the image pair. The SRC method first extracted the local neighborhood patches from the reference image to construct a local dictionary, then reconstructed the background of the current image by sparse representation on the extracted dictionary. Finally, change areas were obtained through background subtracting. The illumination variations between two images were corrected automatically by the representation coefficients, and SRC method based on patches can filter local contrast effectively to detect changing areas. A combination of SRC and some other change detection methods can improve the accuracy of the detection result. In the experiments of the passage, for a simulated image pair with small lesions, the AUC and mAP values were 0.985 and 0.864 respectively. For a clinical image pair with a big lesion, the AUC and mAP values of the combination of SRC and iterative robust homomorphic surface fitting (IRHSF) were 0.989 and 0.969 respectively. Experimental results showed that SRC was more robust than RPCA for the illumination variations and could detect the changing area more effectively than pixel-wised subtraction as it was involved with more neighborhood information.
		2019 Vol. 38 (5): 540-548 [Abstract] ( 373 ) HTML (1 KB) PDF (10213 KB) ( 74 )

	549	Diagnosis of Benign and Malignant Breast Tumors Using a Quantitative Radiomic Method
		Zhao Shuang, Wei Guohui, Ma Zhiqing, Zhao Wenhua
		DOI: 10.3969/j.issn.0258-8021.2019.05.005
		Breast cancer is one of the malignant cancers with the highest mortality rate in women. To improve the diagnostic efficiency and provide more objective and accurate diagnosis results, we used a public data set BreaKHis of pathological images of breast tumors in 82 patients by radiomic method. We extracted grayscale features, Haralick texture features, local binary patterns (LBP) features and Gabor features of 139-dimensional radiomic features of breast tumor pathology images from the data set. The principal component analysis (PCA) was employed to reduce the dimensionality of the omics. After that we constructed a diagnostic model of breast tumors by using four different classifiers including random forest (RF), extreme learning machine (ELM), support vector machine (SVM), k-nearest neighbor (kNN) and evaluated the different feature sets mentioned above. Results showed that the classification of radiomics features based on support vector machine was the best. The accuracy rate reached 88.2%, the sensitivity reached 86.62%, and the specificity reached 89.82%. The proposed method provided a new detection solution for the prediction of benign and malignant breast tumors, which would greatly improve the accuracy of clinical diagnosis of benign and malignant breast tumors.
		2019 Vol. 38 (5): 549-557 [Abstract] ( 465 ) HTML (1 KB) PDF (3453 KB) ( 312 )

	558	Image-Guided Arthroscopic Surgery Based on Virtual Endoscopic Technology
		Cui Xiwen, Chen Fang, Han Boxuan, Ma Cong, Ma Longfei, Liao Hongen
		DOI: 10.3969/j.issn.0258-8021.2019.05.006
		Knee arthroscopic surgery is a kind of minimally invasive surgery using endoscope, which is associated with small incisions and a shorter hospital stay. While in the meantime, arthroscopic surgery uses two dimensional image as visual guidance, thus brings the drawback of lacking depth information, getting blocked easily, counting on surgeon′s experience. Our research brought up an arthroscopic surgery guiding system based on virtual-vision technology with pre-operation imaging. Through this system, pre-operation image collaborated with intraoperative image, and thus made the surgery procedure more convenient and reliable. Virtual-vision rendering method was used for arthroscopic surgery guidance. The calibration procedure for the arthroscope and the registration procedure for the whole system were finished at first. Then the target of virtual rendering for the pre-operation imaging in corresponding viewing position was achieved, during which, real-time tracking for the arthroscope was essential. We set up the arthroscopic surgery guiding system based on external tracking information and pre-operation imaging. Software interface for the system was developed with computer vision rendering algorithm. A knee model experiment was conducted to validate the system, where the mean square error (MSE) for the calibration of tracking device was within 1 mm,and the MSE for the image fusion was below 0.7 mm, both the separate and fusion image for guidance was provided in real-time. The surgery guiding system described in this research could provide enhanced image guidance for knee arthroscopic surgery. Multi-source imaging was made good use in our system, which would provide the surgeons with direct and precise surgery guiding information.
		2019 Vol. 38 (5): 558-565 [Abstract] ( 519 ) HTML (1 KB) PDF (4825 KB) ( 687 )

	566	Research on Effects of Transcranial Direct Current Stimulation on EEG in Children with Autism Spectrum Disorder
		Wen Fang, Pang Jiao, Li Xiaoli, Kang Jiannan
		DOI: 10.3969/j.issn.0258-8021.2019.05.007
		The autism spectrum disorder (ASD) is a complex developmental disorder characterized by impairments of social communication and repetitive behaviors. Transcranial direct current stimulation (tDCS) is a non-invasive brain stimulation technique that can be conducted on many neuropsychiatric and neurological disorders such as epilepsy, Parkinson’s disease, Alzheimer’s disease, major depression and schizophrenia. It is safe and easy-to-use, because thousands of tDCS sessions have been reported without any serious adverse effects. In this study, we enrolled 24 ASD children who received transcranial direct current stimulation (tDCS) brain modulation. Among those, 12 ASD children received 10 treatments over the dorsolateral prefrontal cortex (DLPFC) twice a week and the other 12 ASD children received sham-stimulation twice a week as controls. Power spectrum and multiscale entropy algorithm were used to evaluate the effects. Results showed that 4~8 Hz theta band decreased significantly (P<0.05) in the whole brain before and after intervention, the frontal lobe decreased from (1.13±0.07) dB/Hz to (0.96±0.06)dB/Hz, from (1.18±0.05) dB/Hz to (1.03±0.07)dB/Hz in the left temporal lobe, from (1.43±0.06) dB/Hz to (1.16±0.03)dB/Hz in the central region, from (1.14±0.09) dB/Hz to (0.96±0.04)dB/Hz in the right temporal lobe, and from (1.39±0.06) dB/Hz to (1.09±0.03)dB/Hz in the occipital lobe. Entropy values in parietal lobe (P3, Pz, C3, C4), occipital lobe (O1) and DLPFC (F3) were significantly increased by calculating the entropy values at 15 scales after intervention. The tDCS had positive effects on the autistic children and it was safe and non-invasive, therefore it would be helpful for the rehabilitation of autistic children.
		2019 Vol. 38 (5): 566-572 [Abstract] ( 527 ) HTML (1 KB) PDF (1783 KB) ( 754 )

	573	Design of Portable Wireless Intelligent Interactive System Based on Eye Movement Signal
		Gao Dongrui, Wang Rungui, Ying Shaofei, Jiang Dong, Chen Jiaxin, Zong Xin, Dong Lijuan, Song Xiaoyu, Wang Lutao
		DOI: 10.3969/j.issn.0258-8021.2019.05.008
		Patients with motor impairment are unable to control modern electronic devices (mobile phones, iPAD, etc.) to communicate with the outside world. It is difficult for them to integrate into information-based society. In this paper, a portable wireless intelligent interactive system was designed to help users control electronic devices with their own eye movement signals to help them communicate with the outside world. The system was made up of analog and digital circuits. The analog circuit was used for filtering and amplification. The digital circuit was used to convert analog signals to digital signals, and then the digital signals were analyzed and processed in real time. It included using short-term energy endpoints to obtain starting and ending positions of signal activity segments, selecting location, amplitude, and wave peak number as signal characteristics, and then using zero-crossing analysis and dynamic threshold method to identify different types of eye movement signals. At last a state machine was used to define the corresponding relationship between different types of eye signals and mouse movements in the electronic devices. Fifteen subjects were enrolled for the test, they were required to control five different eye movements according to random instructions, including looking up, down, left, right and active blinking, and each action was performed 100 times in total. Results showed that the average accuracy of each eye movement was more than 94%, and the minimum average information transfer rate was more than 22 bits/min. In conclusion, this system could use the eye movement to replace the mouse to control electronic devices, realizing functions of character input, making phone calls, listening to music, and browsing the web.
		2019 Vol. 38 (5): 573-580 [Abstract] ( 464 ) HTML (1 KB) PDF (980 KB) ( 573 )

	581	Physiological Control of Rotary Blood Pumps Based on BP Neural Networks
		Zhu Zhuoling, Zhao Weiguo, Huang Feng
		DOI: 10.3969/j.issn.0258-8021.2019.05.009
		In this study, a physiological controller of rotary blood pumps based on BP neural networks was proposed, which realized the adaptive adjustment of the controller while the states of the recipients are changing. Choosing the mean arterial pressure as the control object, the controller adopted a three-layer neural network to optimize the PID parameters of the blood pump controller online when the physiological state of the circulatory system changed. This method was verified numerically on the mathematical model of blood circulation system. The controller was carried out under different conditions including left ventricle failure, physiological changing in systemic resistance and dynamic changing in left ventricular contractility. Results showed that in all of the cases the BP neural network based controller could overcome the disturbances well and the mean arterial pressure was stabilized at 100 mmHg after about 150 s since the controller took effect, with the steady-state error of 0 mmHg. This control method could adapt to the changes of various physiological states of the circulatory system and provide an effective control method of rotary blood pump for the subsequent in vitro and animal experiments.
		2019 Vol. 38 (5): 581-589 [Abstract] ( 468 ) HTML (1 KB) PDF (1120 KB) ( 371 )

	590	Construction and Simulation of Three-Layer EIT Model in Gastric
		Li Zhangyong, Liu Zhaoyu, Ran Peng, Xiang Shangzhi, Ma Chengqun, Wang Wei
		DOI: 10.3969/j.issn.0258-8021.2019.05.010
		The aim of this work is to establish a three-layer electrical impedance tomography (EIT) model to study the regularity of gastric emptying under different electrode models and the change of gastric impedance measurement signals when the electrical conductivity of gastric contents is different. In this paper, COMSOL MULTIPHYSICS simulation software was used to build a human abdominal model and a three-layer electrode model with 16 electrodes on each layer. The excitation was simulated by a 5 mA excitation current, and the relative and adjacent electrode model was set. The simulation experiments were carried out under the condition of insulation test meal of 0.054 S/m, conductive test meal of 1 S/m and neutral test meal of 0.5 S/m, and the stomach volume changed from 2 times to 1 time. Simulations were performed according to the pattern of relative excitation and adjacent excitation. The data of the measurement voltage were analyzed, and the voltage sensitivity δ and the boundary voltage measurement dynamic rangeU~ were used to evaluate the stability of the measurement voltage and the detection effect. The larger the value of δ, the better the system detection effect was. The relative electrode model had higher measurement voltage sensitivity δ, and it was smaller than that of the conductive test meal during the insulation test. The voltage sensitivity δ of the adjacent model was larger than that of the conductive test meal during the insulation test, and the dynamic range of the boundary measurement is also relatively larger. When the conductivity was 0.054 S/m, the results of two incentive modes were 34.13 and 34.25, respectively; when the conductivity was 1 S/m, the two incentive modes were 33.60 and 26.68, respectively. The three-layer EIT model could provide 24×23×3 groups of horizontal measurement data sets, as well as the scalable cross-stimulation data sets according to the requirements, providing more information about gastric emptying and effectively reflecting the situation of gastric contents and the information relationship of gastric emptying process.
		2019 Vol. 38 (5): 590-598 [Abstract] ( 475 ) HTML (1 KB) PDF (8258 KB) ( 133 )

Reviews

	599	Application and Research Development of Spatial Filtering Method in Brain-Computer Interfaces
		Wang Tao, Ke Yufeng, Wang Ningci, Liu Wentao, An Xingwei, Ming Dong
		DOI: 10.3969/j.issn.0258-8021.2019.05.011
		With the development of brain-computer interface (BCI) technology, it has become a hot research topic to improve the classification accuracy of BCI system by optimizing data processing algorithms. In recent years, in a variety of BCI paradigms, spatial filtering has been found to improve the signal-to-noise ratio of EEG signals and optimize feature extraction, thus improving the classification accuracy of BCIs. Therefore, spatial filtering has an important application value in EEG-based BCIs. This paper reviewed the commonly used spatial filtering methods for EEG signal processing in BCIs, including basic spatial filters, spatial filters for EEG preprocessing and feature dimensionality reduction, and spatial filters used for feature extraction in BCIs based on induced rhythms and evoked potentials. Finally, we summarized the current problems and future developments of spatial filtering methods for EEG in BCIs.
		2019 Vol. 38 (5): 599-608 [Abstract] ( 593 ) HTML (1 KB) PDF (1834 KB) ( 656 )

	609	The Research Progress of Transcranial Direct Current Stimulation in the Treatment of Alzheimer′s Disease
		Luo Yinpei, Li Nian, Wen Huizhong, Tian Xuelong
		DOI: 10.3969/j.issn.0258-8021.2019.05.012
		Alzheimer′s disease (AD) is an irreversible neurodegenerative disease that seriously affects the physical and mental health and daily activities of patients. It is currently not possible to effectively prevent the course of AD through drug and cognitive training. Transcranial direct current stimulation (tDCS) is a non-invasive neuromodulation technique that is safe, economical, and well tolerated. It has been widely used in many studies of neurodegenerative diseases in recent years. Studies have found out that tDCS can improve learning and memory and cognitive dysfunction in AD patients by regulating synaptic accessibility, participating in inflammation regulation and regulating local blood flow. Therefore, tDCS is a very promising treatment for AD. In order to better understand and promote the application of tDCS in AD, this paper reviewed the application status and mechanisms of tDCS in AD.
		2019 Vol. 38 (5): 609-620 [Abstract] ( 544 ) HTML (1 KB) PDF (902 KB) ( 718 )

	621	Research Progress of the Effects of Body Temperature on Brain Cognitive Function and Regulation Strategies
		Zhou Peng, Zhou Linying, An Xingwei, Yang Jiajia, Wan Baikun, Ming Dong
		DOI: 10.3969/j.issn.0258-8021.2019.05.013
		The long-term exposure to extreme conditions of high or low temperature would make body uncomfortable, which greatly affects the brain cognitive function and may impair the alertness, attention, working memory and other functions. Skin temperature stimulation, as a simple physical method of body temperature regulation, can stabilize and improve the cognitive function of the brain to a certain extent. This paper reviewed the current research status and existing challenges. It was showed that there has not been clear understanding of the mechanism of regulating body temperature to protect brain nerves and improve cognitive function. It is still necessary to explore the optimal target stimulus temperature and time to achieve the goal of rapid and effective improvement of brain cognitive ability. Looking forward to future research strategies, we proposed that it should closely combine the mechanism of hypothermic stimulation on brain nerve protection and the intrinsic mechanism of brain cognitive function to study and formulate temperature control strategies, so as to figure out the best scheme of fast and effective regulation, as well as safe and stable protection of brain cognitive function.
		2019 Vol. 38 (5): 621-627 [Abstract] ( 510 ) HTML (1 KB) PDF (764 KB) ( 705 )

	628	Research Progress of Stereolithographic 3D Printing of Soft Tissue Engineering Scaffolds
		Xu Kehui, Li Jiaojiao, Li Xiangyu, Chen Jialong
		DOI: 10.3969/j.issn.0258-8021.2019.05.014
		Stereolithographic 3D printing technology has characteristics of fast forming speed and high precision, which allow it to accurately control the size, shape and strength of 3D printing soft tissue scaffolds, complete the high matching customization, and effectively solve the huge gap of the soft tissue replacement. At present, the application scope of this technology depends on the properties of photosensitive materials. Firstly, it is necessary to have appropriate viscosities, curing times and curing shrinkage rates for accurately controlling the soft tissue scaffolds by stereolithographic 3D printing. Secondly, printing tissues need to meet the mechanical properties (such as strength, hardness, toughness) and good biocompatibility (such as promoting cell adhesion, proliferation and differentiation), which were directly affected by degradation, porosity and vascularization. This review discussed the performance requirements and the improvement methods and the trends of photosensitive materials, aiming to provide guidance and insights to the development of photosensitive printing materials for soft tissue engineering.
		2019 Vol. 38 (5): 628-635 [Abstract] ( 553 ) HTML (1 KB) PDF (1869 KB) ( 731 )

Communications

	636	Study on the Preparation and Properties of Fabrics of Stent-Grafts in-situ Fenestration
		Wang Shaoxia, Lin Jing, Lao Jihong, Wang Lu
		DOI: 10.3969/j.issn.0258-8021.2019.05.015

		2019 Vol. 38 (5): 636-640 [Abstract] ( 450 ) HTML (1 KB) PDF (2991 KB) ( 407 )

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