Chinese Journal of Biomedical Engineering

Home

Editorial Board

Download

Quick Search

Office Online

Current Issue

Download Articles

Download

Links

2024 Vol. 43, No. 5 Published: 2024-10-20

Reviews Communications Regular Papers Expert Consensus

Expert Consensus

	513	Expert Consensus on Technical Specification of Human Liver-on-Chip for Evaluating Drug-Induced Liver Injury
		Compiling Expert Group for Expert Consensus on Technical Specification of Human Liver-on-Chip for EvaluatingDrug-Induced Liver Injury
		DOI: 10.3969/j.issn.0258-8021.2024.05.001
		Drug-induced liver injury (DILI) is an important part of drug safety evaluation. Existing evaluation methods mainly rely on in vitro two-dimensional cell culture models and animal models, which cannot simulate the human complex physiological environment and have limited predictive ability for drug metabolism and toxicity. Therefore, it is necessary to develop in vitro liver replacement models that can accurately reproduce normal functions of human liver and its responses to drug toxicity. Human liver-on-chips not only have more comprehensive physiological microstructures and functions than two-dimensional cell models, but also eliminate species differences between animals and humans, and more effectively simulate the physiological or pathological state of human liver. At present, the human liver-on-chip, as a novel in vitro anthropomorphic model for DILI detection, is still in the initial development stage and has not yet formed a unified standard or guideline. However, its enormous potential and application prospects have attracted widespread attention. This consensus is oriented to the evaluation of DILI, relying on the human liver-on-chip platform, combined with relevant domestic and foreign model construction schemes, toxicity detection indicators and drug lists, as well as research literatures and data, to form this expert consensus and provide relevant suggestions. The consensus is aiming to promote the development of human liver-on-chip technology and its application in the detection of DILI.
		2024 Vol. 43 (5): 513-524 [Abstract] ( 81 ) HTML (1 KB) PDF (889 KB) ( 101 )

Regular Papers

	525	Retinal Lesion Segmentation with Retinopathy Guidance Deterministic Representation from Diffusion Models
		Xie Yingpeng, Qu Junlong, Xie Hai, Wang Tianfu, Lei Baiying
		DOI: 10.3969/j.issn.0258-8021.2024.05.002
		Acquiring a comprehensive segmentation map of retinal lesions is a crucial step in developing an automated, interpretable diagnostic tool for retinopathy. However, the inherent diversity and complexity of retinal lesions, coupled with the high cost of precise annotation, pose substantial challenges to traditional supervised learning approaches. Recent advances suggest that representation learning can mitigate the reliance on extensive annotated data by pre-training robust image representation models on large-scale unlabeled datasets. In this study, we introduced an innovative representation learning framework based on denoising Diffusion Probabilistic Models (DDPM), specifically tailored to capture the subtle and localized variations in medical imagery, thereby providing precise feature representations for the segmentation of retinal lesions. Utilizing unlabeled fundus images, our approach learnt the reverse process of Markov diffusion, establishing a foundation for extracting pixel-level representations. A retinal lesion grading classifier, informed by domain knowledge of retinopathy severity and lesion correlation, was implemented to guide the reverse diffusion process to enhance representations pertinent to lesions. The guided representations served as a repository of intrinsic semantic information, offering robust image representations for downstream retinal segmentation tasks. In experiments on multiple fundus image datasets, our method achieved average Dice coefficients of 0.872 for optic cup and 0.877 for optic disc segmentation with only 50 samples. For diabetic retinopathy lesions, it reached a Dice coefficient of 0.664, and for age-related macular degeneration lesions, 0.513, demonstrating diffusion-based representation’s generality and effectiveness across various complex retinal conditions.
		2024 Vol. 43 (5): 525-538 [Abstract] ( 55 ) HTML (1 KB) PDF (12656 KB) ( 40 )

	539	Study on Extraction Method of Reduced-Dimensionality Ensemble Empirical Mode Decomposition for Cardiopulmonary Signals Based on Electrical Impedance Tomography Technology
		Li Kun, Li Weichen, Guo Yitong, Wang Weice, Wang Yu, Yan Xiaoheng, Shi Xuetao
		DOI: 10.3969/j.issn.0258-8021.2024.05.003
		Real-time acquisition of cardiac ejection and pulmonary ventilation activity information is of great clinical significance. To separate both cardiac ejection and pulmonary ventilation activity signals simultaneously from chest electrical impedance tomography (EIT) data, this study proposed a novel signal extraction method named reduced-dimensionality ensemble empirical mode decomposition (RDEEMD) method. A total of 9 volunteers were recruited for EIT chest data collection. Firstly, this method classified the measurement channels based on the strength of the cardiac activity signal of the chest EIT data under breath holding state. Subsequently, the ensemble empirical mode decomposition method was used to decompose the EIT data under autonomous breathing state, and the decomposed components were categorized based on spectral characteristics to obtain the lung ventilation EIT signal. Combined with the band-pass filtering method and based on the aforementioned channel classification, the heart activity EIT signal was obtained by reduced-dimensionality of the heart activity signal. Finally, the EIT image sequences of the ventilation phase and cardiac phase were reconstructed. The results showed that the highest power spectral peak for lung ventilation (52.71 ± 1.39) dB in the lung area of the ventilation phase image can be obtained through RDEEMD, the highest power spectral peak for cardiac activity (43.05 ± 3.26) dB in the heart area of the cardiac phase image can be obtained through RDEEMD, indicating a fine reservation of ventilation and cardiac activity information. Meanwhile, the power spectral peak related to cardiac activity in the heart area of the ventilation phase image obtained by RDEEMD (10.02 ± 2.65) dB is the lowest among these methods, indicating that the effect of cardiac activity was well inhibited, compared to the reference method, there were significant differences (P<0.05). These results showed that the method RDEEMD could effectively separate signals related to lung ventilation and heart activity, preserving respective activity information and suppressing the influence of the heart in lung imaging. Finally, it can effectively suppress interference signal and lay the foundation for providing more accurate treatment strategy guidance in clinical practice.
		2024 Vol. 43 (5): 539-549 [Abstract] ( 47 ) HTML (1 KB) PDF (3174 KB) ( 33 )

	550	Cross-Subjects EEG Emotion Recognition Based on Scaled Convolutional Attention Network
		Chen Binbin, Wu Tao, Chen Lifei
		DOI: 10.3969/j.issn.0258-8021.2024.05.004
		Emotion recognition based on EEG signals has become a medical aid to emotion regulation intervention because it can objectively reflect human physiological and psychological states. To address the problem of poor generalization performance of emotion recognition caused by existing methods ignoring the differences in channel data distribution between individuals, a new cross-subjects emotion recognition method based on scaled convolutional attention network was proposed in this work. Based on the extraction of emotion quantification features in multi-channel electroencephalography (EEG) signals, a novel scaled convolutional attention network was constructed to establish the synergistic change relationship of emotional features of different channels and scales, and the weight of the synergistic relationship was automatically learned by model training, and finally the domain invariant representation of the emotion polarity was obtained, which improved the generalization performance of emotion recognition across multiple individuals EEG. The emotion EEG dataset SEED and SEED-IV were used to identify emotions cross-subjects with100 665 and 100 950 EEG samples, and the recognition accuracy of the proposed method was 89.63% and 75.65% in the three and four-class of emotions. Particularly, the robustness of the proposed model outperformed most of the existing methods when there were changes in the number of individuals. The results showed that the proposed method was able to effectively extract the domain-invariant representation of emotional polarity.
		2024 Vol. 43 (5): 550-560 [Abstract] ( 50 ) HTML (1 KB) PDF (7534 KB) ( 11 )

	561	Functional Connectivity Analysis of Cortical Muscles for Stroke Patients Based on Partial Transfer Entropy
		Shi Zhengyi, Xie Qiurong, Wang Xiaoling, Li Yurong
		DOI: 10.3969/j.issn.0258-8021.2024.05.005
		Motor dysfunction is the main symptom after stroke, which is generally thought to be caused by the damage of neural network that controls motor function. To explore the neuromuscular control mechanism of stroke patients, in this paper, the functional cortical muscular coupling (FCMC) is used as a tool to collect the electroencephalogram (EEG) signals of 13 stroke patients and 13 healthy controls during the extension movement. Meanwhile, the Electromyogram (EMG) signals of triceps brachii, anterior deltoid toe, middle toe, back toe, biceps brachii, pectoralis major and trapezius muscle are obtained by FCMC. Moreover, by locating the cerebral cortex source, the raw EEG signal is obtained, afterwards the active brain region corresponding to the action is also determined by clustering, finally the cortical muscle functional connection is obtained by using partial transfer entropy. The functional connectivity of ICsF with AD, ICsB with BIC, ICsC with PD, PM, and UT in the upstream channels was significantly enhanced in stroke patients compared with healthy controls (For example, the healthycontrol group is 0.033±0.031, the stroke patient is 0.092±0.083, P<0.05). The functional connectivity of BIC with ICsA, ICsB and ICsC, TRI and UT with ICsC in the downstream channels was significantly enhanced in stroke patients compared with healthy controls (For example, the healthy control group is 0.113±0.092, and the stroke patients are 0.198±0.105, P<0.05). Ipsilateral cortical muscle functional connectivity is present in stroke patients. The experimental results show that this paper explores the cortical muscle functional connection effect after stroke from a new angle, and proves that there exists ipsilateral cortical muscle functional connection in stroke patients, which further effectively promotes the understanding of neuromuscular coupling mechanism after stroke.
		2024 Vol. 43 (5): 561-570 [Abstract] ( 47 ) HTML (1 KB) PDF (3674 KB) ( 25 )

	571	Emotional Face Fixation Strategies in High and Low Autistic Trait
		Liu Shuang, Wang Dandan, Xue Huiqin, Zhang Ludan, Guan Xin, Liu Wei, Ming Dong
		DOI: 10.3969/j.issn.0258-8021.2024.05.006
		Autistic traits refers to a personality trait that is continuously distributed and widely present in normal people, and has similar behavioral and cognitive characteristics to autism spectrum disorders. Thehigh autistic trait (HAT) has certain defects in the ability to recognize emotional faces, leading to insufficient social cognitive and social communication abilities in this group. This study focused on emotional face gaze strategies in groups with HAT and low autistic traits (LAT) and explained their abnormal social performance from the perspective of eye movement. We designed an emotional face recognition paradigm covering four basic emotions including happiness, neutrality, anger, and sadness. The autism spectrum quotient and eye movement data were collected from 23 subjects with HAT and 20 subjects with LAT. Using traditional visual analysis methods based on areas of interest and heat map tools, features such as emotional face gaze heat map, the eye movement entropy, and spatial scanning similarity were extracted to explore the gaze processing strategies of HAT and LAT groups when scanning emotional faces. The results of emotional face gaze heat map showed, compare with the LAT group, the HAT group focused more on the nose and mouth, and less on the eyes, indicating that there was a phenomenon of gaze avoidance towards the eyes in the HAT group. The eye movement entropy of HAT and LAT was the lowest at 0° of face angle and the highest at 90° of face angle (16.364±0.390 vs 16.706±0.387, P<0.05). All results of the similarity of LAT were higher than that of HAT, but there was no statistical significance. And HAT group was difficult to learn in the process of emotional face recognition to the law, lack of strategic leading to observation. In summary, this study compared the differences in eye movement characteristics between HAT and LAT groups when performing emotional recognition tasks, enriching the research on eye movement processing strategies for emotional faces in this group, and exploring the causes of social disorders and cognitive dysfunction in high loneliness trait groups from the perspective of eye movement.
		2024 Vol. 43 (5): 571-581 [Abstract] ( 38 ) HTML (1 KB) PDF (6871 KB) ( 21 )

	582	An Observation Method of Human-Computer Interaction for Identification of Target MoleculesBased on U-Net Convolutional Neural Network
		Zhang Xinfeng, Yin Wenbin, Fang Jinpeng, Zhang Xinmei
		DOI: 10.3969/j.issn.0258-8021.2024.05.007
		The observation of molecules that play an important role in life activities is an important way to discover intrinsic mechanisms of the life activities. Most of existing biomedical image processing methods focus on the detection and identification of specific substances, however, it is difficult to adapt to changing demands of scientific research. To this end, this paper proposed a human-computer interaction method based on U-Net convolutional neural network to identify all the same molecules in biomedical images, such as nucleic cell, proteins, etc. First, the U-Net convolutional network was used to convert molecular images to deep feature maps, and then the features of the target molecules were used to match on the entire feature map to detect all the same molecules of interest. Then, the CSR-DCF (discriminative correlation filter with channel and spatial reliability) algorithm was used to build a multi-target tracker to achieve continuous tracking of the target molecules. Experimental results showed that the proposed method was able to quickly detect similar molecules of interest through simple human-computer interaction, and obtain important information on the number, distribution and interactions of target molecules. Attention-based U-Net and U-Net performed consistently on 200 static test images randomly selected from Nucleus, Human Protein Atlas, Bacteria and Blood Red Cell datasets, withaverage precision mean values of 0.912 5 and 0.898 1, respectively. At the same time, the tracking of targets in the dynamic images of mouse stem cells was accurate and stable, proving the effectiveness of the method to meet the needs of microscopic life process observation in life science research.
		2024 Vol. 43 (5): 582-595 [Abstract] ( 35 ) HTML (1 KB) PDF (12269 KB) ( 11 )

Reviews

	596	A Review on Intelligent Detection Methods of Gastrointestinal Tract Lesions Based on WirelessCapsule Endoscopy Images
		Fan Shanhui, Wei Shangguang, Wei Kaihua, Fan Yihong, Lv Bin, Fan Kai, Li Lihua
		DOI: 10.3969/j.issn.0258-8021.2024.05.008
		Wireless capsule endoscopy (WCE) is a non-invasive and painless method for gastrointestinal (GI) tract examination, but it produce tens of thousands of images during the examination, which may only contain a few abnormal images. The review work highly relies on the clinical experience of physicians, which is time-consuming and easily cause false detection and/or missed detection. Therefore, exploring automated detection methods of WCE abnormal images to aid to clinical diagnosis with high accuracy and efficiency has become a research hotspot in this field. This paper reviewed recent researches on intelligently detecting GI lesions on WCE images. We first introduced the basic principles and evaluation metrics of intelligent detection methods, and then outlined the researches on intelligent detection methods of WCE abnormal images in recent years from the perspectives of traditional machine learning and deep learning algorithms, and summarized the advantages and shortcomings of the reported methods. Finally, this paper discussed and concluded the current challenges and possible future research directions in automatic lesion classification of WCE.
		2024 Vol. 43 (5): 596-608 [Abstract] ( 51 ) HTML (1 KB) PDF (927 KB) ( 65 )

	609	Development Status of General Brain-Computer Interface Software Platform
		He Feng, Wang Changhao, Wang Kun, Mei Jie, Luo Ruixin, Xu Minpeng, Ming Dong
		DOI: 10.3969/j.issn.0258-8021.2024.05.009
		Brain-computer interface (BCI) is a next-generation disruptive human-machine interaction technology that establishes a direct information pathway between the brain and external devices. Building a complete BCI system requires both hardware devices and software systems. The software system includes steps such as stimulus presentation, signal acquisition, signal processing, and command feedback, which have high technical requirements. To address this challenge, researchers have released a series of general BCI software platforms to provide solutions for BCI software development. The article focused on the stages of BCI technology development and systematically outlined key features of general BCI software platforms at different stages. Furthermore, we elaborated on the functional characteristics of 11 representative products. At the end of this review, we discussed the challenges and possible solutions of existing general BCI software platforms, aiming to promote continuous optimization and development of these platforms for practical applications.
		2024 Vol. 43 (5): 609-619 [Abstract] ( 45 ) HTML (1 KB) PDF (1738 KB) ( 80 )

	620	Research Progress of Biological 3D Printing Strategy in Three-Dimensional Tissue Constructionof Functionalized Skeletal Muscle
		Zhang Feihu, Li Ting, Lin Zhiwei, Li Yanbing, Huang Wenhua
		DOI: 10.3969/j.issn.0258-8021.2024.05.010
		Skeletal muscle is essential for body movement, and loss of motor function due to volumetric muscle loss (VML) can limit the ability to move. Transplant surgery is the main clinical approach, however, is still limited by the morbidity of the donor site, the lack of donor tissue, and the need for a highly skilled surgical team. Biological 3D printing technology provides a more efficient and accurate method for the manufacture of functional muscles, and a feasible method for the regeneration and repair of VML muscles. This paper reviewed the research status of biological 3D printing technology in the construction of functional skeletal muscle tissue structure, as well as research progress of biological 3D printing and biological ink in functional muscle structures, mainly focusing on the construction of vascularized and neuralized skeletal muscle biomimetic tissue. The effects of various functional skeletal muscle structures constructed by 3D printing on the repair of skeletal muscle defects were summarized. Finally, development trends and existing limitations in the field were proposed.
		2024 Vol. 43 (5): 620-630 [Abstract] ( 55 ) HTML (1 KB) PDF (5356 KB) ( 83 )

Communications

	631	Multi-Task Information Extraction of Electronic Medical Records Based on HierarchicalParameter Transformation
		Gan Weinan, Xia Xiaoqin, Ouyang Xiaoping, Yang Jianmin
		DOI: 10.3969/j.issn.0258-8021.2024.05.011

		2024 Vol. 43 (5): 631-635 [Abstract] ( 39 ) HTML (1 KB) PDF (803 KB) ( 14 )

	636	Clinical Effect Analysis of Laser-Assisted Irrigation in Endodontic Revascularization of Young Permanent Teeth
		Dai Li, Song Yumeng, Wei Zheng, Zhang Wang, Yuan Qin, Miao Leiying, Xing Xianghui
		DOI: 10.3969/j.issn.0258-8021.2024.05.012

		2024 Vol. 43 (5): 636-640 [Abstract] ( 31 ) HTML (1 KB) PDF (4345 KB) ( 24 )

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