Chinese Journal of Biomedical Engineering

Home

Editorial Board

Download

Quick Search

Office Online

Current Issue

Download Articles

Download

Links

2025 Vol. 44, No. 1 Published: 2025-02-20

Reviews Communications Regular Papers

Regular Papers

	1	Research on Hybrid Evoked Paradigms Based on Weak Spatial Modulation Visual Evoked Potentials
		Zhou Xiaoyu, Xiao Xiaolin, Xu Minpeng, Ming Dong
		DOI: 10.3969/j.issn.0258-8021.2025.01.001
		The spatial modulation characteristics of visual evoked potentials (VEPs) provide an effective approach for designing user-friendly and practical brain-computer interface systems. However, the spatially modulated VEPs are usually weak and have low signal-to-noise ratios, making it crucial to study how to efficiently elicit and accurately identify weak spatially modulated VEPs. In this study, we employed visual stimuli with a radius smaller than 0.5° of visual angle and designed two mixed spatial modulation induction paradigms including the "transient stimuli serial induction paradigm" and the "steady-state and transient stimuli parallel induction paradigm". Twelve healthy subjects participated in the experiments, and the two paradigms were quantitatively compared by calculating the spatial modulated signal-to-noise ratio (sm-SNR) and offline classification accuracy. Results indicated that the average sm-SNR of the "Low-Frequency Transient" spatial modulation feature under the "transient stimuli serial induction paradigm" could reach 0.0148, significantly higher than that under the "steady-state and transient stimuli parallel induction paradigm" for the "low-frequency transient", "high-frequency steady-state", and their mixed spatial modulation features. The offline classification results were consistent with the feature analysis results, showing that the spatial modulation VEPs recognition accuracy was higher under the "transient stimuli serial induction paradigm", with an average classification accuracy of 85%. This study is expected to provide a reference for the design of high-performance visual brain-computer interfaces based on the spatial modulation characteristics of VEPs.
		2025 Vol. 44 (1): 1-10 [Abstract] ( 51 ) HTML (1 KB) PDF (7772 KB) ( 32 )

	11	Visually Induced Emotional Valence Evaluation using Physiologic Network-Based Brain-Heart Interaction
		Cai Zhipeng, Gao Hongxiang, Li Jianqing, Liu Chengyu
		DOI: 10.3969/j.issn.0258-8021.2025.01.002
		In the past decade, complex physiological interactions between brain and heart during emotional processing have become a research hotspot. This study employed network physiological methods to analyze the time-delay stability(TDS) quantization indices between electroencephalogram (EEG) and electrocardiogram (ECG) signals recorded during visually induced emotional stimuli in 414 data entries from the Dreamer database, exploring the brain-heart interaction under visual emotional stimulation. The research revealed asymmetric connectivity between the brain hemispheres during emotional processing, particularly highlighting the dominant role of the right hemisphere in these interactions. EEG analysis emphasized the critical role of low-frequency bands (δ, θ, α) in the transmission of emotional information, with δ-θ coupling in the frontal region being particularly crucial for emotional regulation. In high-valence emotional states, the %TDS value of δ-θ coupling (0.78±0.05) was significantly higher than in low-valence states (0.65±0.04). Additionally, the average link strength of brain-heart interactions in low-valence states reached its peak (0.68±0.06), while in high-valence states, it decreased to the lowest (0.59±0.03). These findings not only enhance our understanding of the synchronization mechanisms between the cortical brain and heart in emotional processing but also enrich the knowledge base in neurophysiology and emotional science.
		2025 Vol. 44 (1): 11-20 [Abstract] ( 49 ) HTML (1 KB) PDF (16131 KB) ( 23 )

	21	Classification of Dysphonia in Parkinson′s Disease Based on Hierarchical Fractional Spectrogram
		Xue Zaifa, Lu Huibin, Lin Liqin, Zhang Tao
		DOI: 10.3969/j.issn.0258-8021.2025.01.003
		Dysphonia is one of the early symptoms of Parkinson′s disease. Most of the existing deep learning-based classifications of dysphonia in Parkinson′s disease are based on spectrogram and convolutional neural network, but both of them suffer from deficiencies such as single angle and restricted receptive field, respectively, which lead to insufficient information extraction. This paper proposed a classification method for Parkinson′s disease based on hierarchical fractional spectrogram. Firstly, by adding angle rotation factors, the dysphonia signal was transformed into the fractional spectrogram to enhance the ability of extracting energy information from different angles. Then the parameters of the Swin Transformer network pre-trained on ImageNet were transferred and fine-tuned to solve the problem of small data size. Finally, the combination of hierarchical structure and shifted window-based self-attention mechanism expanded the receptive field and realized multi-scale information fusion, which effectively improved the Parkinson′s disease classification accuracy. The results on Database-1 (240 samples collected by the Department of Neurology of Medicine, Istanbul University) and Database-2 (1 404 samples collected by Tangshan Workers′ Hospital and Kailuan Mental Health Center) showed good stability of the proposed method and achieved accuracy of 97.80 % and 98.75 % on the two datasets, respectively, with better performance than all compared advanced methods. Our proposed method provides a new perspective for analyzing articulation disorders in Parkinson′s disease.
		2025 Vol. 44 (1): 21-33 [Abstract] ( 25 ) HTML (1 KB) PDF (4870 KB) ( 26 )

	34	Vessel Segmentation in Coronary Angiography Images Based on Deep Convolution and Multi-Level Scale Feature Fusion
		Xu Yang, Zhai Nannan, Ni Weizhen, Tan Qiang, Wang Jinjia
		DOI: 10.3969/j.issn.0258-8021.2025.01.004
		Coronary angiography is a significant diagnostic and therapeutic modality for coronary heart disease and other cardiovascular diseases. The accurate and expeditious segmentation of blood vessels is of paramount importance to the diagnosis and treatment of cardiovascular diseases. Existing coronary angiography vessel segmentation algorithms have been shown to have several shortcomings, including a weak segmentation ability for fine vessels, poor connectivity of segmented vessels, and a lack of resistance to noise and artefacts. This study proposes an enhanced U-shape segmentation network, termed HAM-UNet.UNet, which utilises the advantages of the Transformer structure′s long-distance dependence and cross-domain hopping connectivity. The proposed methods include contextual hierarchical aggregation and multiscale feature fusion. Firstly, a series of image preprocessing methods are employed to enhance certain features of the original coronary angiography images and expand the experimental data. Then, the preprocessed images are segmented by the HAM-UNet method.The encoder combines both deep convolution and residual structure, which can efficiently capture global features and effectively enhance the network detail perception, thus improving the segmentation accuracy while increasing the segmentation connectivity. The decoder performs multi-scale feature fusion and up-sampling hopping connections, improving the global perception of the network and reducing the influence of irrelevant information.The datasets used are from 221 images from the General Hospital of Tianjin Medical University and 494 images from the First Hospital of Qinhuangdao City. On both datasets, the HAM-UNet algorithm achieves an accuracy of 0.983 and 0.998, respectively. As demonstrated in Figure 8, the IOUs are 0.857 and 0.908, and the Dice scores are 0.842 and 0.883, respectively. This indicates that the overall segmentation performance is superior to that of U-Net, Att-UNet and other algorithms.
		2025 Vol. 44 (1): 34-42 [Abstract] ( 45 ) HTML (1 KB) PDF (7449 KB) ( 30 )

	43	Multiscale Attention-Based CNN Model for Vessel Enhancement of Coronary Angiography
		Zhou Peng, Wang Guangpu, Gao Hui, Qin Zewei, Wang Shuo, Yu Hui
		DOI: 10.3969/j.issn.0258-8021.2025.01.005
		Coronary angiography records the dynamic process of developing vessels with blood flow, which is interfered by cardiac motion, resulting in poor image quality, seriously affecting the diagnosis of physicians, and at the same time is not conducive to the intelligent assisted diagnosis of coronary heart disease. In this paper, a vessel enhancement network for coronary angiography with multi-scale attention based on CNN was proposed. The network consisted of multi-scale attention block (MAB) and large kernel attention tail (LKAT). The MAB contained multi-scale large kernel attention (MLKA) and gated spatial attention block (GSAB), the module was able to extract more local and global vessel information while avoiding the grid effect. The LKAT showed the ability to aggregate long-range information, which improved the characterization of reconstructed vascular features and thus enhanced the reconstruction quality of coronary angiography images. The 2 666 coronary datasets in the experiment were manually labeled by medical expert pairs, and the obtained vessel segmentation labels were used as masks, which were superimposed onto the Gaussian-filtered preprocessed images as coronary enhancement labels. Compared with the existing state-of-the-art methods, the enhancement effect was remarkable, with PSNR and SSIM reaching 34.880 1 and 0.973 2, respectively. Moreover, the enhanced segmentation results achieved Dice and IoU of 0.851 4 and 0.741 3, respectively, with Acc and Recall reaching 98.55% and 89.05%. The experimental results showed that the method realizes intelligent enhancement of coronary angiography images, and it also facilitates the follow-up of the intelligent auxiliary diagnosis of coronary heart disease processing.
		2025 Vol. 44 (1): 43-51 [Abstract] ( 45 ) HTML (1 KB) PDF (7298 KB) ( 16 )

	52	Glandular Segmentation Based on Shape Stream and Multi-Scale Feature Fusion
		Lin Jiawen, Chen Susu, Lin Zhiming, Li Li, Weng Qian
		DOI: 10.3969/j.issn.0258-8021.2025.01.006
		Meibomian gland imaging technology is widely used in the classification diagnosis, management and personalized treatment of dry eye syndrome. Direct observation and qualitative evaluation by ophthalmologists may result in low subjective and reproducible evaluation. To improve the diagnostic efficiency of ophthalmologists, researchers have proposed a series of gland segmentation method for infrared meibomian gland images based on U-Net. However, the segmentation results are still not ideal at image edges, at locations of reflective points, and in areas with dense glandular structures. Considering the characteristics of infrared meibomian gland imaging and glandular distribution, this paper proposed a glandular segmentation model SS-UNet based on shape stream and multi-scale feature fusion, introduced an atrous convolution module to enhance the model′s feature extraction ability, designed a shape stream auxiliary branch to fully learn the shape information of glands, used a multi-scale feature fusion module to obtain feature representations of glands with different thicknesses. To verify the validity of the model, a fully annotated dataset containing 203 infrared meibormian gland images were collected by the ophthalmology department of Fujian Provincial Hospital and used to conduct comparative experiments with other advanced medical segmentation models in the same experimental environment, perform module ablation analysis, and display the visualization results. The experimental results showed that the Acc, Dice, IoU indicators of SS-UNet reached 94.62%, 80.94%, and 68.17%, respectively, which were improved by 0.36%, 1.41%, and 1.95% compared to the benchmark network U-Net, significantly improving the gland segmentation results. This work has shown that SS-UNet was able to fully utilize information such as the shape and scale of glands to solve incorrect segmentation problems such as glandular adhesions and missed detections, effectively improving segmentation accuracy and providing objective basis for assisting clinical diagnosis.
		2025 Vol. 44 (1): 52-65 [Abstract] ( 31 ) HTML (1 KB) PDF (9148 KB) ( 16 )

	66	Fast White Blood Cell Detection Algorithm Based on Lightweight Network
		Chen Liang, Guo Huihui, Yin Tao
		DOI: 10.3969/j.issn.0258-8021.2025.01.007
		Due to the large variety and morphological differences of white blood cells, and overlap, adhesion, cell boundary blurring and color change in blood microscopy, the traditional system based on image detection has difficulty in feature extraction, poor detection accuracy and insufficient stability. To address these problems, a white blood cell rapid detection algorithm based on lightweight network structure was proposed. Firstly, the algorithm used MobileNetv3 as the feature extraction network, and proposed a dual-channel pyramid feature fusion structure TCPF-Net to complete the feature fusion for its insufficient feature extraction ability. The algorithm improved the feature extraction ability of white blood cell images with blur, color change and different shapes. After that, the algorithm abandoned the large target detection head of the detection network and only retained the small and medium target detection head for the special aspect ratio and scale characteristics of white blood cells, which improved the detection speed of the algorithm for the white blood cells. Finally, the algorithm used the intersection over union parameter when the complete anchor frame overlaped with the target to complete the optimization of the regression loss function of the detection network position, and improved the detection ability of the algorithm for overlapping and adherent cells. The experiment was conducted using 40x microscopic images of human blood stained with the Romanowsky staining method. With the validation of 8,848 white blood cell images, the meanaverage precision (mAP) of the lightweight network algorithm for white blood cell detection reached 98.8%, representing 1.1% improvement compared to the original network. Simultaneously, theframes per second (FPS) reached 54.19, indicating a 32% increase compared to the original network, achieving the rapid and precise detection of white blood cells.
		2025 Vol. 44 (1): 66-76 [Abstract] ( 27 ) HTML (1 KB) PDF (15394 KB) ( 5 )

	77	Cognitive Assessment of Brain Connectivity Features Based on fNIRS
		Tian Yizhu, Zhang Ye, Qin Tian, Mao Zhenfang, Li Deyu, Xia Meiyun
		DOI: 10.3969/j.issn.0258-8021.2025.01.008
		There are sex differences in cognitive decline in Alzheimer′s disease (AD), and whether this sex difference is present in the pre-AD or even earlier stages is unknown. In this study, functional near-infrared spectroscopy (fNIRS) technology was used to compare the sex differences in whole brain activity in the verbal fluency test (VFT) between the healthy group and the MCI group, aiming to find fNIRS functional connectivity features significantly related to gender and behavior, followed by finding cognitive assessment indicators based on functional connectivity, laying a foundation for carrying out sex-specific AD early screening. The study found significant differences between healthy subjects and MCI subjects in the main effects of sex, and the differential functional connections were located in the long-range functional connectivity between the middle and back parts of the brain and the hemispheres. Compared with the healthy group, the connectivity difference pattern between males and females in the MCI group changed, and the difference of functional connectivity z-values decreased. Among them, in the VFT task, functional connectivity z-values of the parietal lobe (MCI males: P=0.03; MCI females: P=0.02), especially the parietal-left inferior parietal (MCI females: MMSE, P=0.01; correct word number, P=0.05) were significantly correlated with gender and behavioral performance, which may be potential features for early evaluation of females with AD and deserve more attention in the future.
		2025 Vol. 44 (1): 77-88 [Abstract] ( 29 ) HTML (1 KB) PDF (6618 KB) ( 23 )

Reviews

	89	Advances in Microfluidic Sorting of Extracellular Vesicles
		Cao Changming, Li Zhen, Tian Yanhong, An Rong, Ren Tianling
		DOI: 10.3969/j.issn.0258-8021.2025.01.009
		Extracellular vesicles (EVs) are one type of membrane-enclosed particles released by cells into the extracellular environment. EVs are ubiquitously present in various body fluids including blood and urine. They serve as biomarkers for numerous diseases, owing to their content of diverse biomolecules like proteins and nucleic acids. The extraction and analysis of EVs are crucial for the rapid diagnosis and treatment of diseases. Traditional techniques employed for sorting EVs, such as differential ultracentrifugation, need a substantial amount of sample material and involve costly instrumentation. Conversely, microfluidic sorting techniques offer several benefits, including miniaturization, high recovery rates and integration, thus making them more suitable for the application in medical institutions and clinics. This review categorized microfluidic sorting methods into label-free passive, label-free active, fixed-substrate immunoaffinity and free-bead immunoaffinity sorting, introduced the research progress of these methods in the field of EVs isolation. The characteristics including purity, recovery, flux and difficulty in chip fabrication of each method are summarized and compared in the review. In addition, the review envisaged the future development directions for microfluidic sorting of EVs, which included sorting indexes improvement, sorting costs reduction and separation cut-off size decrement. For label-free sorting, acoustophoresis is more promising due to the advantages in achieving high purity and throughput, as well as the capability to sort nano-sized EVs. Acoustophoresis also enables monolithic integration of acoustic sensing modules for EV concentration. For immunoaffinity sorting, the technology using free beads has advantages of high capture efficiency and chip reusability, which is suitable for continuous and batch sorting scenarios.
		2025 Vol. 44 (1): 89-100 [Abstract] ( 29 ) HTML (1 KB) PDF (10705 KB) ( 11 )

	101	Application and Progress of Deep Learning in Circulating Tumor Cell Detection
		Zhu Shuai, Liu Ming, Yang Jianbo, He Defeng, Zhao Ming
		DOI: 10.3969/j.issn.0258-8021.2025.01.010
		Circulating tumor cells (CTCs) as typical biomarkers in liquid biopsy bear great promise in early diagnosis, prognosis judgment and therapeutic monitoring of tumor. CTCs are scarce, diverse and heterogeneous in peripheral blood, and their detection tasks face multiple challenges such as low accuracy and poor specificity. Deep learning has been widely used in biomedical research and clinical applications. Efficient and accurate automated CTCs detection using deep learning models has become a new research hotspot. In thisarticle, we systematically reviewed the related research work of deep learning applied to CTCs detection in recent years. Existing method theory, key technologies, and performance analysis were described in CTCs detection process, including sample preparation, data acquisition and preprocessing, and construction of deep learning models. At last, the current challenges and future trends of deep learning in CTCs detection were discussed.
		2025 Vol. 44 (1): 101-111 [Abstract] ( 41 ) HTML (1 KB) PDF (4192 KB) ( 31 )

	112	Research Progress on Application of Polyurethane and its Modified Materials forSpinal Implants
		Xiao Ruxin, Liao Liqiong, Song Jian
		DOI: 10.3969/j.issn.0258-8021.2025.01.011
		Polyurethane (PU) and its modified materials have been investigated for uses in clinical practice as emerging bone biomaterials due to their excellent physical and chemical properties and good biocompatibility. In clinical practice, the composition and processing methods of PU affect its mechanical properties, which in turn affects the stability of PU implants. By combing the effects of different synthesis and processing methods of PU materials, the biomechanical and clinical trial effects of their application in spinal implants were analyzed. Investigations of PU application in spinal implants were summarized, and the shortcomings of PU in the application were analyzed as well. In addition, development directions of PU in the application of spinal implants were discussed.
		2025 Vol. 44 (1): 112-123 [Abstract] ( 37 ) HTML (1 KB) PDF (3106 KB) ( 26 )

Communications

	124	Lightweight EMG Artifact Detection Method Based on Improved YOLO Model for EEG
		Sun Ge, Lin Weihong, Lou Hongwei, Han Jinbo
		DOI: 10.3969/j.issn.0258-8021.2025.01.012

		2025 Vol. 44 (1): 124-128 [Abstract] ( 29 ) HTML (1 KB) PDF (1530 KB) ( 22 )

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