Pub Date : 2022-11-05DOI: 10.1109/CISP-BMEI56279.2022.9979842
Zhengyang Liu, Xiaoyu Song, Shiwei Zheng, Liming Li
Retinal prostheses aim to restore functional vison for the blind suffering from retinal degeneration diseases. The visual acuity recovered by a retinal prosthesis is often restricted by the limited electrode number and density. Virtual channel (VC) is a strategy for improving resolution without increasing number of physical electrodes. Some previous studies have investigated the characteristics of traditional stactic VC without considering the effects of time intervals when multiple VCs are involved. However, the actual performance of VCs will be largely influenced by the time intervals between time adjacent VCs. In this study, we analyzed the effect of time intervals between two consecutive VCs on the activation pattern of computational model of retinal ganglion cells (RGCs), which we define as the dynamic virtual channel (DVC) condition. We found that the activated RGC area is quite different at varied time intervals. In addition, the activated RGC area can be kept consistent after optimizing intervals between VC pairs. Our results indicate that performance of DVC can be well optimized by appropriate stimulating parameter selection.
{"title":"A Preliminary Exploration on Dynamic Virtual Channel for Epiretinal Stimulation: A Modelling Study","authors":"Zhengyang Liu, Xiaoyu Song, Shiwei Zheng, Liming Li","doi":"10.1109/CISP-BMEI56279.2022.9979842","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9979842","url":null,"abstract":"Retinal prostheses aim to restore functional vison for the blind suffering from retinal degeneration diseases. The visual acuity recovered by a retinal prosthesis is often restricted by the limited electrode number and density. Virtual channel (VC) is a strategy for improving resolution without increasing number of physical electrodes. Some previous studies have investigated the characteristics of traditional stactic VC without considering the effects of time intervals when multiple VCs are involved. However, the actual performance of VCs will be largely influenced by the time intervals between time adjacent VCs. In this study, we analyzed the effect of time intervals between two consecutive VCs on the activation pattern of computational model of retinal ganglion cells (RGCs), which we define as the dynamic virtual channel (DVC) condition. We found that the activated RGC area is quite different at varied time intervals. In addition, the activated RGC area can be kept consistent after optimizing intervals between VC pairs. Our results indicate that performance of DVC can be well optimized by appropriate stimulating parameter selection.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131484693","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-05DOI: 10.1109/CISP-BMEI56279.2022.9980251
Shan Liu, Xiaoqing Wu
Nowadays, online public opinion is developing rapidly and has a great influence. If it is not guided in time, it will easily endanger the stability of society. Therefore, this paper focused on building a network public opinion dissemination model, and analyzed the evolution law of public opinion dissemination at different stages, so as to help the government do a good job of public opinion prevention and intervention to build a good network environment. This paper took the hot event of “the Fukushima nuclear wastewater” as an example, drawn the evolution curve of public opinion by fitting the collected data, and divided the development of public opinion into three stages, and researched their network structure and public opinion diffusion law respectively. Research shows that the process of public opinion dissemination can be divided into three stages: incubation period, diffusion period, and recession period. In different stages, the network structure has different characteristics.
{"title":"Public Opinion Evolution and Communication Stages in Complex Network","authors":"Shan Liu, Xiaoqing Wu","doi":"10.1109/CISP-BMEI56279.2022.9980251","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9980251","url":null,"abstract":"Nowadays, online public opinion is developing rapidly and has a great influence. If it is not guided in time, it will easily endanger the stability of society. Therefore, this paper focused on building a network public opinion dissemination model, and analyzed the evolution law of public opinion dissemination at different stages, so as to help the government do a good job of public opinion prevention and intervention to build a good network environment. This paper took the hot event of “the Fukushima nuclear wastewater” as an example, drawn the evolution curve of public opinion by fitting the collected data, and divided the development of public opinion into three stages, and researched their network structure and public opinion diffusion law respectively. Research shows that the process of public opinion dissemination can be divided into three stages: incubation period, diffusion period, and recession period. In different stages, the network structure has different characteristics.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128243514","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-05DOI: 10.1109/CISP-BMEI56279.2022.9979984
Hao Chen, Wenming Chen, Li Chen, Xiong Yang, Xin Ma
Minimum foot clearance (MFC) is defined as the minimum vertical distance between the lowest point of the swing foot or shoe and the walking surface in gait, which is now considered as a critical gait parameter for predicting trip-related fall risks. Different MFC methods have been used to assess fall risks by tripping, with the analytical method proposed by Begg et al. being the most widely used one. Since this method is based on assumption of a 2D triangular geometric model of the foot, the effects of out-of-plane rotations of the foot/shoe on MFC were not completely known. Furthermore, the accuracy of the MFC maybe influenced by factors such as shoe type, limiting its potential applications in clinical scenarios. Thus, this study proposes a novel method to calculate MFC parameter (called 3D-MFC) based on 3D modeling of the “virtual” markers of the shoe. By using a dynamic point-tracking technology, the 3D-MFC can automatically extract the MFC height while subject walking. From the Bland-Altman analysis, it was shown the 3D-MFC agreed well with that of the Begg's 2D-Geometric method. However, the mean absolute error (MSE) and root mean square error (RMSE) of the 3D-MFC method were less than 1 mm, which significantly outperformed the 2D-Geometric method, especially for subjects using rocker-bottom shoe. It is suggested that the 3D-MFC has potential to be an effective solution for identifying the MFC parameters and is expected to be used for biomechanical assessment of trip-related fall risks in the elderly.
{"title":"3D-MFC: A method for computing critical gait parameters based on virtual marker tracking","authors":"Hao Chen, Wenming Chen, Li Chen, Xiong Yang, Xin Ma","doi":"10.1109/CISP-BMEI56279.2022.9979984","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9979984","url":null,"abstract":"Minimum foot clearance (MFC) is defined as the minimum vertical distance between the lowest point of the swing foot or shoe and the walking surface in gait, which is now considered as a critical gait parameter for predicting trip-related fall risks. Different MFC methods have been used to assess fall risks by tripping, with the analytical method proposed by Begg et al. being the most widely used one. Since this method is based on assumption of a 2D triangular geometric model of the foot, the effects of out-of-plane rotations of the foot/shoe on MFC were not completely known. Furthermore, the accuracy of the MFC maybe influenced by factors such as shoe type, limiting its potential applications in clinical scenarios. Thus, this study proposes a novel method to calculate MFC parameter (called 3D-MFC) based on 3D modeling of the “virtual” markers of the shoe. By using a dynamic point-tracking technology, the 3D-MFC can automatically extract the MFC height while subject walking. From the Bland-Altman analysis, it was shown the 3D-MFC agreed well with that of the Begg's 2D-Geometric method. However, the mean absolute error (MSE) and root mean square error (RMSE) of the 3D-MFC method were less than 1 mm, which significantly outperformed the 2D-Geometric method, especially for subjects using rocker-bottom shoe. It is suggested that the 3D-MFC has potential to be an effective solution for identifying the MFC parameters and is expected to be used for biomechanical assessment of trip-related fall risks in the elderly.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"59 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133357348","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-05DOI: 10.1109/CISP-BMEI56279.2022.9979858
Yu Wang, Jiaxi Sun
The traditional action recognition algorithm based on manual feature extraction is relatively complex and has low recognition accuracy. This paper presents a video action recognition algorithm based on double branch convolutional neural network, which includes two separate convolutional neural networks in sequence. The training network effectively extracts spatio-temporal features through 3D convolution and GRU layers. Then, the features extracted from the training network are input into the test network for classification. The accuracy of the proposed algorithm is 95.0% on UCF-101 dataset. By comparing with other benchmark methods, the accuracy and effectiveness of this method are verified.
{"title":"Video Human Action Recognition Algorithm Based on Double Branch 3D-CNN","authors":"Yu Wang, Jiaxi Sun","doi":"10.1109/CISP-BMEI56279.2022.9979858","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9979858","url":null,"abstract":"The traditional action recognition algorithm based on manual feature extraction is relatively complex and has low recognition accuracy. This paper presents a video action recognition algorithm based on double branch convolutional neural network, which includes two separate convolutional neural networks in sequence. The training network effectively extracts spatio-temporal features through 3D convolution and GRU layers. Then, the features extracted from the training network are input into the test network for classification. The accuracy of the proposed algorithm is 95.0% on UCF-101 dataset. By comparing with other benchmark methods, the accuracy and effectiveness of this method are verified.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130256365","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-05DOI: 10.1109/CISP-BMEI56279.2022.9980027
Wenchao Lyu, Yang Song, Guanghong Liu, Kuoye Han, Jupeng Li
To suppress coordinate system offset and improve track correlation precision, a point cloud registration based correlation method for multi-radar was proposed in this paper. Coordinate system offset parameters were estimated by point cloud registration. Then based on the registration parameters, the coordinate system offset was calibrated, all the track points' positions were updated. Finally, the updated tracks were correlated by the K-Nearest Neighbors method. To evaluate the performance of the method, simulation data was used in experiments. Results indicate that the method improves the correlation accuracy and adapts to the coordinate offset in multi-radar systems.
{"title":"Point Cloud Registration based Track Correlation Method","authors":"Wenchao Lyu, Yang Song, Guanghong Liu, Kuoye Han, Jupeng Li","doi":"10.1109/CISP-BMEI56279.2022.9980027","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9980027","url":null,"abstract":"To suppress coordinate system offset and improve track correlation precision, a point cloud registration based correlation method for multi-radar was proposed in this paper. Coordinate system offset parameters were estimated by point cloud registration. Then based on the registration parameters, the coordinate system offset was calibrated, all the track points' positions were updated. Finally, the updated tracks were correlated by the K-Nearest Neighbors method. To evaluate the performance of the method, simulation data was used in experiments. Results indicate that the method improves the correlation accuracy and adapts to the coordinate offset in multi-radar systems.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"44 11","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114023340","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-05DOI: 10.1109/CISP-BMEI56279.2022.9979956
Lei Yang, Fanke Meng, Boyuan Yu, Zhan Xu, Yi Gong
The core network is the most significant mobile communication system control unit and is responsible for user registration, authentication, and session establishment. Storing subscriber information is an essential primary function of the core network, and in order to ensure the security and stability of the core network operation, high requirements are put forward for the core network storage unit. Due to its intrinsic nature, blockchain has overwhelming advantages over traditional databases in terms of robustness, data consistency, high availability, etc., and has nearly unlimited scalability. With more and more user devices accessing the network, the service pressure of operators is increasing with each passing day, so we propose a core network solution based on blockchain. The blockchain is used as the data storage unit of the core network, and the smart contract is utilized to read and write the data and control the permissions. The blockchain key pair further refined the data permission management of different network elements. This solution can improve the security and robustness of the core network.
{"title":"Blockchain-based User Data Storage And Protection in 5G Core Network","authors":"Lei Yang, Fanke Meng, Boyuan Yu, Zhan Xu, Yi Gong","doi":"10.1109/CISP-BMEI56279.2022.9979956","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9979956","url":null,"abstract":"The core network is the most significant mobile communication system control unit and is responsible for user registration, authentication, and session establishment. Storing subscriber information is an essential primary function of the core network, and in order to ensure the security and stability of the core network operation, high requirements are put forward for the core network storage unit. Due to its intrinsic nature, blockchain has overwhelming advantages over traditional databases in terms of robustness, data consistency, high availability, etc., and has nearly unlimited scalability. With more and more user devices accessing the network, the service pressure of operators is increasing with each passing day, so we propose a core network solution based on blockchain. The blockchain is used as the data storage unit of the core network, and the smart contract is utilized to read and write the data and control the permissions. The blockchain key pair further refined the data permission management of different network elements. This solution can improve the security and robustness of the core network.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116046087","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-05DOI: 10.1109/CISP-BMEI56279.2022.9980073
Wanxin Cheng, Xiang Chen, Hui Zhang, C. He, Han Yang, Jin Li
Simulation in an individualized real head model based on magnetic resonance imaging (MRI) is increasingly required for safety design of electroconvulsive therapy (ECT), which aims to reduce the potential risk of side-effect. To simplify the complicated operational procedures of conventional methods, this paper designs and implements the integrated simulation software for multi-channel ECT on MATLAB platform. Any step in the simulation procedure, in detail, MRI data read, MRI normalization and segmentation, tissue classification and reconstruction, mesh generation, geographic model generation, electrodes placement, parameter configuration and electric field simulation, which are previously performed on COMSOL, MATLAB, AVIZO and other large-scale software platforms, can be performed by just clicking the corresponding item in a MATLAB user-friendly interface. It uses the MATLAB toolboxes to replace specific platforms such as AVIZO to generate mesh. Although it still takes COMSOL as the core, the simulation computations and other operations on COMSOL run in the background by calling corresponding interface functions. In addition, it provides options of the simulation channel number for future developing ECT technique. The software significantly improves the efficiency of data processing and reduces the time cost, which is an easy-to-use tool for ECT simulation.
{"title":"Design and Implementation of Integrated Simulation Software for Multichannel Modified Electroconvulsive Therapy","authors":"Wanxin Cheng, Xiang Chen, Hui Zhang, C. He, Han Yang, Jin Li","doi":"10.1109/CISP-BMEI56279.2022.9980073","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9980073","url":null,"abstract":"Simulation in an individualized real head model based on magnetic resonance imaging (MRI) is increasingly required for safety design of electroconvulsive therapy (ECT), which aims to reduce the potential risk of side-effect. To simplify the complicated operational procedures of conventional methods, this paper designs and implements the integrated simulation software for multi-channel ECT on MATLAB platform. Any step in the simulation procedure, in detail, MRI data read, MRI normalization and segmentation, tissue classification and reconstruction, mesh generation, geographic model generation, electrodes placement, parameter configuration and electric field simulation, which are previously performed on COMSOL, MATLAB, AVIZO and other large-scale software platforms, can be performed by just clicking the corresponding item in a MATLAB user-friendly interface. It uses the MATLAB toolboxes to replace specific platforms such as AVIZO to generate mesh. Although it still takes COMSOL as the core, the simulation computations and other operations on COMSOL run in the background by calling corresponding interface functions. In addition, it provides options of the simulation channel number for future developing ECT technique. The software significantly improves the efficiency of data processing and reduces the time cost, which is an easy-to-use tool for ECT simulation.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116489819","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-05DOI: 10.1109/CISP-BMEI56279.2022.9979957
Nchongmaje Ndipenoch, A. Miron, Zidong Wang, Yongmin Li
Accurate quantification of retinal Optical Coherence Tomography (OCT) images provides important clinical information of the pathological changes present in age-related macular degeneration (AMD). Currently, monitoring the progress of AMD is mostly performed manually by ophthalmologists, which is time-consuming, difficult and prone to errors. In this work, we have developed a model Deep_ResUNet++ to address this issue and to provide an automatic solution to the problem of simultaneous segmenting retinal layers and fluid regions from retinal OCT images. We have evaluated the method on the Annotated Retinal OCT Images (AROI) dataset. Experimental results demonstrate that our method outperformed the baseline U-Net model, the current state-of-the-art models (UNet_ASPP, ResUnet and ResUnet++) and even the human experts' annotation results, and achieved the best performance by a clear margin with Dice Score above 90% in every single class.
{"title":"Simultaneous Segmentation of Layers and Fluids in Retinal OCT Images","authors":"Nchongmaje Ndipenoch, A. Miron, Zidong Wang, Yongmin Li","doi":"10.1109/CISP-BMEI56279.2022.9979957","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9979957","url":null,"abstract":"Accurate quantification of retinal Optical Coherence Tomography (OCT) images provides important clinical information of the pathological changes present in age-related macular degeneration (AMD). Currently, monitoring the progress of AMD is mostly performed manually by ophthalmologists, which is time-consuming, difficult and prone to errors. In this work, we have developed a model Deep_ResUNet++ to address this issue and to provide an automatic solution to the problem of simultaneous segmenting retinal layers and fluid regions from retinal OCT images. We have evaluated the method on the Annotated Retinal OCT Images (AROI) dataset. Experimental results demonstrate that our method outperformed the baseline U-Net model, the current state-of-the-art models (UNet_ASPP, ResUnet and ResUnet++) and even the human experts' annotation results, and achieved the best performance by a clear margin with Dice Score above 90% in every single class.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114431597","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-05DOI: 10.1109/CISP-BMEI56279.2022.9980167
Xiaoyu Sui, Yankun Cao, Jia Mi, Kemeng Tao, Jing Han, Kun Zhao, Chun Wang, Zhi Liu
Carotid ultrasound is a main and convenient method for diagnosing plaque, Therefore, accurately obtaining plaques information from ultrasound images is essential for further clinical diagnosis. Due to the interference of noise and the differences in artificial technical operations, the missed inspection of the plaques is likely to cause missed inspection. Therefore, a comparative experiment based on the constructing algorithm based on convolutional neural networks is performed to achieve more accurate detection and identification of cervical arterial tube cavity and plaques. First of all, we constructed the carotid artery data set, and then conducted the classification test of the carotid lumen and plaques through the YOLOv5 network based on migration and learning, and used the Faster R-CNN and SSD network for comparison experiments. Experiments show that the average accuracy obtained by YOLOv5 network reaches 0.928 when the IOU value is 0.5, and 0.659 when the IOU value is 0.75, and the average recall rate reaches 0.673, which are higher than the Faster R-CNN and SSD networks; The experiment shows that the average precision of the comprehensive comparison is also better than the other two comparison networks. At the same time, the model calculation speed meets the real-time needs. Therefore, the YOLOv5 network can improve the correctness and practical significance of the detection of the lumen and plaques in terms of carotid image detection.
{"title":"Research on Carotid Ultrasonic Image Detection Based on Convolutional Neural Network","authors":"Xiaoyu Sui, Yankun Cao, Jia Mi, Kemeng Tao, Jing Han, Kun Zhao, Chun Wang, Zhi Liu","doi":"10.1109/CISP-BMEI56279.2022.9980167","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9980167","url":null,"abstract":"Carotid ultrasound is a main and convenient method for diagnosing plaque, Therefore, accurately obtaining plaques information from ultrasound images is essential for further clinical diagnosis. Due to the interference of noise and the differences in artificial technical operations, the missed inspection of the plaques is likely to cause missed inspection. Therefore, a comparative experiment based on the constructing algorithm based on convolutional neural networks is performed to achieve more accurate detection and identification of cervical arterial tube cavity and plaques. First of all, we constructed the carotid artery data set, and then conducted the classification test of the carotid lumen and plaques through the YOLOv5 network based on migration and learning, and used the Faster R-CNN and SSD network for comparison experiments. Experiments show that the average accuracy obtained by YOLOv5 network reaches 0.928 when the IOU value is 0.5, and 0.659 when the IOU value is 0.75, and the average recall rate reaches 0.673, which are higher than the Faster R-CNN and SSD networks; The experiment shows that the average precision of the comprehensive comparison is also better than the other two comparison networks. At the same time, the model calculation speed meets the real-time needs. Therefore, the YOLOv5 network can improve the correctness and practical significance of the detection of the lumen and plaques in terms of carotid image detection.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"157 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114650140","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-11-05DOI: 10.1109/CISP-BMEI56279.2022.9980163
Xin Xu, Jiaxing Zhang
With the rapid development of informati on society, people's demand for personal privacy a nd property protection has become stronger and st ronger. At present, the traditional biometric techno logy has been difficult to meet the needs of social de velopment. Electroencephalography (EEG), as a un ique biometric feature of individuals, has received wide attention from a large number of researchers. In order to solve the problems of difficult to apply in practice and low recognition accuracy due to the induction of specific situations and differences in i ndividual characteristics in EEG data acquisition, a PSO-Attention-RNN (PARNN) recognition mode 1 is proposed in this paper. Firstly, the energy entro py of five rhythms, a-wave, ß-wave, δ-wave, θ-wave and γ-wave, in EEG signals are extracted as featur e vectors by using wavelet packet transform. These features are then input into the PARNN optimized recognition model, and the EEG temporal frequen cy bands corresponding to different emotional mod ules are filtered using particle swarm optimization (PSO), which can lead to the highest recognition ac curacy for the subjects. The whole process was vali dated in a self-collected emotional EEG database. T he results show that the average recognition accura cy of the algorithm in this paper can reach 90.99%, and the recognition accuracy of the positive emotio n module reaches 93.72%.
{"title":"An Identification Recognition Method Based on the Optimization Mechanism of Emotional EEG Module","authors":"Xin Xu, Jiaxing Zhang","doi":"10.1109/CISP-BMEI56279.2022.9980163","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9980163","url":null,"abstract":"With the rapid development of informati on society, people's demand for personal privacy a nd property protection has become stronger and st ronger. At present, the traditional biometric techno logy has been difficult to meet the needs of social de velopment. Electroencephalography (EEG), as a un ique biometric feature of individuals, has received wide attention from a large number of researchers. In order to solve the problems of difficult to apply in practice and low recognition accuracy due to the induction of specific situations and differences in i ndividual characteristics in EEG data acquisition, a PSO-Attention-RNN (PARNN) recognition mode 1 is proposed in this paper. Firstly, the energy entro py of five rhythms, a-wave, ß-wave, δ-wave, θ-wave and γ-wave, in EEG signals are extracted as featur e vectors by using wavelet packet transform. These features are then input into the PARNN optimized recognition model, and the EEG temporal frequen cy bands corresponding to different emotional mod ules are filtered using particle swarm optimization (PSO), which can lead to the highest recognition ac curacy for the subjects. The whole process was vali dated in a self-collected emotional EEG database. T he results show that the average recognition accura cy of the algorithm in this paper can reach 90.99%, and the recognition accuracy of the positive emotio n module reaches 93.72%.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128518950","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：