Depression is a common psychiatric disorder that can lead to depressed moods and even suicidal behavior. Intelligent assessment of depression from multiple physiological and behavioral data and breaking the limitations of traditional methods is the focus of this research area. In this paper, a multi-modal fusion depression assessment model based on attention mechanisms is proposed to predict the severity of depression from visual, acoustic and text modalities. By training and testing on an improved dataset, the proposed multi-modal feature layer fusion model based on attention mechanisms (MFF-Att) is validated to be superior to unimodal prediction models and achieved good results in depression assessment. The root mean square error (RMSE) and mean absolute error (MAE) of the proposed model on the development set are 4.03 and 3.05, respectively, which are better than the baseline and state-of-the-art results.
{"title":"A Multi-modal Feature Layer Fusion Model for Assessment of Depression Based on Attention Mechanisms","authors":"Congcong Wang, Decheng Liu, Kemeng Tao, Xiaoxiao Cui, Gongtang Wang, Yuefeng Zhao, Zhi Liu","doi":"10.1109/CISP-BMEI56279.2022.9979894","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9979894","url":null,"abstract":"Depression is a common psychiatric disorder that can lead to depressed moods and even suicidal behavior. Intelligent assessment of depression from multiple physiological and behavioral data and breaking the limitations of traditional methods is the focus of this research area. In this paper, a multi-modal fusion depression assessment model based on attention mechanisms is proposed to predict the severity of depression from visual, acoustic and text modalities. By training and testing on an improved dataset, the proposed multi-modal feature layer fusion model based on attention mechanisms (MFF-Att) is validated to be superior to unimodal prediction models and achieved good results in depression assessment. The root mean square error (RMSE) and mean absolute error (MAE) of the proposed model on the development set are 4.03 and 3.05, respectively, which are better than the baseline and state-of-the-art results.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"36 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":"122390006","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-BMEI56279.2022.9980053
W. Shao, Jin-Ye Li, Wei-Wei Luo, Mei-Lin Liu, H. Deng
It is known image priors are essential to blind deconvolution. Reweighted graph total variation (RGTV), as a new prior to substitute the most classic TV, is shown superior to TV as well as several other state-of-the-art models in terms of both theoretical and empirical performance. In this paper, we take a step forward providing a simpler geometric view to RGTV, instead of the previous graph spectral interpretation made in the graph frequency domain. In specific, we formulate blind deblurring just via use of a derivative of the Leclerc loss, which is geometrically proved an appropriate candidate to promote the piecewise smoothing and sharpening desired by RGTV. A by-product of such a perspective is to closely relate blind and non-blind deblurring in a fairly naive fashion. A fast algorithm is then deduced to update the sharp image and blur kernel alternately, through implementing our simplified RGTV as a reweighted-L1 regularizer rather than a graph L1-Laplacian regularizer. Numerous experiments on challenging blurred images show a much better performance of the proposed approach than original RGTV, in terms of both effectiveness and efficiency. Additionally, the proposed method achieves a comparable or superior performance to other state-of-the-art methods, either model-based or deep learning-based ones.
众所周知，图像先验是盲反卷积的必要条件。Reweighted graph total variation (RGTV)作为一种替代经典TV的新方法，在理论和实证表现上都优于TV以及其他几种最先进的模型。在本文中，我们向前迈进了一步，为RGTV提供了更简单的几何视图，而不是以前在图频域中进行的图谱解释。具体来说，我们通过使用勒克莱尔损失的导数来制定盲去模糊，这在几何上被证明是促进RGTV所需的分段平滑和锐化的合适候选。这种观点的副产品是以一种相当幼稚的方式将盲法和非盲法去模糊紧密地联系在一起。通过将我们的简化RGTV实现为重加权l1正则化器而不是图l1 -拉普拉斯正则化器，推导出一种快速的算法来交替更新清晰图像和模糊核。在具有挑战性的模糊图像上进行的大量实验表明，所提出的方法在有效性和效率方面都比原始RGTV有更好的性能。此外，所提出的方法实现了与其他最先进的方法(基于模型或基于深度学习的方法)相当或更好的性能。
{"title":"A Geometric View to Reweighted Graph Total Variation Blind Deconvoluton: Making It Faster and Better","authors":"W. Shao, Jin-Ye Li, Wei-Wei Luo, Mei-Lin Liu, H. Deng","doi":"10.1109/CISP-BMEI56279.2022.9980053","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9980053","url":null,"abstract":"It is known image priors are essential to blind deconvolution. Reweighted graph total variation (RGTV), as a new prior to substitute the most classic TV, is shown superior to TV as well as several other state-of-the-art models in terms of both theoretical and empirical performance. In this paper, we take a step forward providing a simpler geometric view to RGTV, instead of the previous graph spectral interpretation made in the graph frequency domain. In specific, we formulate blind deblurring just via use of a derivative of the Leclerc loss, which is geometrically proved an appropriate candidate to promote the piecewise smoothing and sharpening desired by RGTV. A by-product of such a perspective is to closely relate blind and non-blind deblurring in a fairly naive fashion. A fast algorithm is then deduced to update the sharp image and blur kernel alternately, through implementing our simplified RGTV as a reweighted-L1 regularizer rather than a graph L1-Laplacian regularizer. Numerous experiments on challenging blurred images show a much better performance of the proposed approach than original RGTV, in terms of both effectiveness and efficiency. Additionally, the proposed method achieves a comparable or superior performance to other state-of-the-art methods, either model-based or deep learning-based ones.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"112 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":"122940254","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-BMEI56279.2022.9980063
Li Feng, Mei Yu, Yang Song, Ruitao Chen, Liuyan Cao, G. Jiang
Stereoscopic omnidirectional videos (SOVs) can provide users with continuous immersive visual experience, but the distortions introduced in the process of its processing, coding transmission, visualization bring great challenges to its quality assessment. In this paper, a blind SOV quality assessment method including spatial perception model (SPM) and temporal perception model (TPM) is designed based on the spatio-temporal separability of neurons in the region 18 of visual cortex, and the characteristics of stereoscopic and omnidirectional perception are considered. In the SPM, the fractal dimension is introduced to design the pixel domain feature extraction module of single viewpoint; based on the theory of binocular summation and binocular difference to regulate binocular behavior, the two-stage gain control and maximum response model are combined to construct the binocular perception model, and the orientation feature in transform domain of the binocular perception map is extracted to achieve the complementary role of content perception of single viewpoint and binocular. For the TPM, considering the motion perception of the middle temporal region and the correlation between the left and right viewpoints, a binary statistical model for temporal information extraction is constructed to assist SPM to form functional complementarity. Experimental results show that the proposed method has good quality assessment performance, and has a better consistency with human visual perception.
{"title":"Visual Perception Based Blind Stereoscopic Omnidirectional Video Quality Assessment","authors":"Li Feng, Mei Yu, Yang Song, Ruitao Chen, Liuyan Cao, G. Jiang","doi":"10.1109/CISP-BMEI56279.2022.9980063","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9980063","url":null,"abstract":"Stereoscopic omnidirectional videos (SOVs) can provide users with continuous immersive visual experience, but the distortions introduced in the process of its processing, coding transmission, visualization bring great challenges to its quality assessment. In this paper, a blind SOV quality assessment method including spatial perception model (SPM) and temporal perception model (TPM) is designed based on the spatio-temporal separability of neurons in the region 18 of visual cortex, and the characteristics of stereoscopic and omnidirectional perception are considered. In the SPM, the fractal dimension is introduced to design the pixel domain feature extraction module of single viewpoint; based on the theory of binocular summation and binocular difference to regulate binocular behavior, the two-stage gain control and maximum response model are combined to construct the binocular perception model, and the orientation feature in transform domain of the binocular perception map is extracted to achieve the complementary role of content perception of single viewpoint and binocular. For the TPM, considering the motion perception of the middle temporal region and the correlation between the left and right viewpoints, a binary statistical model for temporal information extraction is constructed to assist SPM to form functional complementarity. Experimental results show that the proposed method has good quality assessment performance, and has a better consistency with human visual perception.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"21 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":"128993421","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-BMEI56279.2022.9980130
Yuzhe Wang, Qiang Du, Li Ke, Yunfeng Bai
It is of great significance that bioimpedance detection for the analysis of cognitive activity in brain functional area. And it is also important for brain science research and clinical diagnosis. In this paper, the cerebral blood flow signal is detected and studied by bioimpedance technique. Firstly, the detection method of brain bioimpedance signal is designed, and the bioimpedance signal of subjects is collected and denoised. Secondly, the characteristic parameters of CBF in bioimpedance and bioimpedance differential signals, including peak value, curve area and comprehensive parameters, are extracted to characterize the activity changes of brain functional areas. Finally, the effectiveness of the detection method is verified by the significant changes of brain bioimpedance characteristic parameters in the prefrontal functional area under the letter memory experiment. The changes of brain impedance characteristic parameters in different periods of memory task are further analyzed. The results show that the bioimpedance characteristic parameters of the prefrontal area are changed by memory task, and the peak bioimpedance and differential bioimpedance are most significantly changed. Therefore, cerebral blood flow bioimpedance technology can monitor brain activity in real time, and with the increase of memory load, cerebral blood flow in the prefrontal functional area will increase correspondingly, cerebral blood flow velocity will also increase correspondingly, and the activation degree of this functional area will be enhanced.
{"title":"Bioimpedance detection and analysis for the prefrontal functional area","authors":"Yuzhe Wang, Qiang Du, Li Ke, Yunfeng Bai","doi":"10.1109/CISP-BMEI56279.2022.9980130","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9980130","url":null,"abstract":"It is of great significance that bioimpedance detection for the analysis of cognitive activity in brain functional area. And it is also important for brain science research and clinical diagnosis. In this paper, the cerebral blood flow signal is detected and studied by bioimpedance technique. Firstly, the detection method of brain bioimpedance signal is designed, and the bioimpedance signal of subjects is collected and denoised. Secondly, the characteristic parameters of CBF in bioimpedance and bioimpedance differential signals, including peak value, curve area and comprehensive parameters, are extracted to characterize the activity changes of brain functional areas. Finally, the effectiveness of the detection method is verified by the significant changes of brain bioimpedance characteristic parameters in the prefrontal functional area under the letter memory experiment. The changes of brain impedance characteristic parameters in different periods of memory task are further analyzed. The results show that the bioimpedance characteristic parameters of the prefrontal area are changed by memory task, and the peak bioimpedance and differential bioimpedance are most significantly changed. Therefore, cerebral blood flow bioimpedance technology can monitor brain activity in real time, and with the increase of memory load, cerebral blood flow in the prefrontal functional area will increase correspondingly, cerebral blood flow velocity will also increase correspondingly, and the activation degree of this functional area will be enhanced.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"4 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":"115234500","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-BMEI56279.2022.9979839
Dan Liu, Li Ke, Qiang Du, Wanni Zu
Magnetic particle imaging (MPI) is an imaging technique used to determine the spatial concentration distribution of superparamagnetic nanoparticles. Tikhonov regularization algorithm is a commonly used reconstruction algorithm in MPI, but the reconstruction accuracy of this method is low, especially when the concentration distribution of magnetic nanoparticles in the image region is widely different, its image quality is difficult to meet the imaging requirements of particle spatial concentration distribution. In this paper, a two-step regularized magnetic particle imaging algorithm is proposed. Firstly, the signal of high concentration particles is extracted and the Tikhonov reconstruction is performed in the first step to obtain the distribution image of high concentration particles. Then, the second step of Tikhonov reconstruction was performed to obtain the low-concentration particle distribution image. Finally, high and low concentration particle distribution images are fused to achieve high quality image of particle concentration distribution. The simulation results show that the maximum concentration ratio of the two samples in MPI is increased by 16 times, and the signal to artifact (SAR) ratio is increased by 16 times. Therefore, the proposed two-step regularization reconstruction algorithm has a good reconstruction effect for magnetic particle imaging with large concentration difference distribution.
{"title":"A two-step regularization reconstruction algorithm for magnetic particle imaging","authors":"Dan Liu, Li Ke, Qiang Du, Wanni Zu","doi":"10.1109/CISP-BMEI56279.2022.9979839","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9979839","url":null,"abstract":"Magnetic particle imaging (MPI) is an imaging technique used to determine the spatial concentration distribution of superparamagnetic nanoparticles. Tikhonov regularization algorithm is a commonly used reconstruction algorithm in MPI, but the reconstruction accuracy of this method is low, especially when the concentration distribution of magnetic nanoparticles in the image region is widely different, its image quality is difficult to meet the imaging requirements of particle spatial concentration distribution. In this paper, a two-step regularized magnetic particle imaging algorithm is proposed. Firstly, the signal of high concentration particles is extracted and the Tikhonov reconstruction is performed in the first step to obtain the distribution image of high concentration particles. Then, the second step of Tikhonov reconstruction was performed to obtain the low-concentration particle distribution image. Finally, high and low concentration particle distribution images are fused to achieve high quality image of particle concentration distribution. The simulation results show that the maximum concentration ratio of the two samples in MPI is increased by 16 times, and the signal to artifact (SAR) ratio is increased by 16 times. Therefore, the proposed two-step regularization reconstruction algorithm has a good reconstruction effect for magnetic particle imaging with large concentration difference distribution.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"2 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":"126888481","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-BMEI56279.2022.9979896
Jun Yin, Meiqi Zhan, Zhaowei Zhang, Lei Wang, Deng-yin Zhang, Xin Xiao
We consider the issue of effective content management in mobile edge caching networks due to the characters of mobile user devices' mobility and heterogeneity. We propose a hierarchical framework for content sharing application in mobile edge caching networks, which mainly includes physical layer and content sharing application layer. The function of the physical layer is to manage mobile users and their content management functions. This layer adopts a clustering method, and the small base stations (stationary edge nodes) distributed in the geographical area are used as cluster heads to manage the mobile users in the area under its jurisdiction, and manage content resources through distributed hash table; the main function of the application layer is to manage the content sharing process between mobile users. To solve the unstable content transmission performance caused by frequent movement of user nodes between regions, we propose a publication-subscription driven content discovery and control scheme, an incentive mechanism for content sharing and a dynamic content provider selection algorithm. Experimental results verify the effectiveness of the proposed hierarchical architecture.
{"title":"Research on the Content Sharing System for Mobile Edge Caching Networks: a Hierarchical Architecture","authors":"Jun Yin, Meiqi Zhan, Zhaowei Zhang, Lei Wang, Deng-yin Zhang, Xin Xiao","doi":"10.1109/CISP-BMEI56279.2022.9979896","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9979896","url":null,"abstract":"We consider the issue of effective content management in mobile edge caching networks due to the characters of mobile user devices' mobility and heterogeneity. We propose a hierarchical framework for content sharing application in mobile edge caching networks, which mainly includes physical layer and content sharing application layer. The function of the physical layer is to manage mobile users and their content management functions. This layer adopts a clustering method, and the small base stations (stationary edge nodes) distributed in the geographical area are used as cluster heads to manage the mobile users in the area under its jurisdiction, and manage content resources through distributed hash table; the main function of the application layer is to manage the content sharing process between mobile users. To solve the unstable content transmission performance caused by frequent movement of user nodes between regions, we propose a publication-subscription driven content discovery and control scheme, an incentive mechanism for content sharing and a dynamic content provider selection algorithm. Experimental results verify the effectiveness of the proposed hierarchical architecture.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"55 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":"126746448","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-BMEI56279.2022.9980138
Zhenhao Zhu, Yuanhui Yu, Yan Wang
If there is no network communication in the Android world, then the Android world is a collection of isolated islands. The most common network request framework in Android development is the Retrofit framework. In addition, Volley and OkHttp, which are widely used, have their own advantages and disadvantages. The best comprehensive performance is the Retrofit framework, which has a convenient and easy-to-use request interface and can easily automate entity resolution. Decouple the parsing and asynchronous request frameworks, and the superior data access and storage performance makes network requests concise and elegant. Finally, an example is given to illustrate the application of Retrofit framework in cloud notes in classes.
{"title":"An Application of Network Communication Technology in Classroom Cloud Notes","authors":"Zhenhao Zhu, Yuanhui Yu, Yan Wang","doi":"10.1109/CISP-BMEI56279.2022.9980138","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9980138","url":null,"abstract":"If there is no network communication in the Android world, then the Android world is a collection of isolated islands. The most common network request framework in Android development is the Retrofit framework. In addition, Volley and OkHttp, which are widely used, have their own advantages and disadvantages. The best comprehensive performance is the Retrofit framework, which has a convenient and easy-to-use request interface and can easily automate entity resolution. Decouple the parsing and asynchronous request frameworks, and the superior data access and storage performance makes network requests concise and elegant. Finally, an example is given to illustrate the application of Retrofit framework in cloud notes in classes.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"4 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":"121889354","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}
Computer-aided segmentation technology is important for clinical treatment of brain tumors. In recent years, U-shaped networks have become mainstream for medical image segmentation, significantly improving the performance of brain tumor segmentation tasks. Since merits of the U -shaped architecture, we propose a new shuffle attention residual U-Net, i.e., SAResU-Net, for brain tumor segmentation application. SAResU-Net combines several shuffle attention (SA) blocks and residual modules with a basic 3D U-Net, where SA blocks are added to skip connection positions to capture the local spatial and channel information. In addition, a self-ensemble module is leveraged to further boost the model performance. Evaluation experimental results on the 2019 and 2020 Brain Tumor Segmentation (BraTS) datasets show that our SAResU-Net is superior to its baseline, especially on the tumor core segmentation task. Moreover, our model achieves DSC values of 79.17%, 90.02% and 82.00% for the enhancing tumor (ET), the whole tumor (WT), and tumor core(TC) on the BraTS 2020 validation dataset, respectively, while on the validation dataset of BraTS 2019, the values are 77.74%, 90.40% and 83.58%, respectively, proving its effectiveness in the application of brain tumor segmentation.
{"title":"SAResU-Net: Shuffle attention residual U-Net for brain tumor segmentation","authors":"Yuqing Zhang, Yutong Han, Dongwei Liu, Jianxin Zhang","doi":"10.1109/CISP-BMEI56279.2022.9979978","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9979978","url":null,"abstract":"Computer-aided segmentation technology is important for clinical treatment of brain tumors. In recent years, U-shaped networks have become mainstream for medical image segmentation, significantly improving the performance of brain tumor segmentation tasks. Since merits of the U -shaped architecture, we propose a new shuffle attention residual U-Net, i.e., SAResU-Net, for brain tumor segmentation application. SAResU-Net combines several shuffle attention (SA) blocks and residual modules with a basic 3D U-Net, where SA blocks are added to skip connection positions to capture the local spatial and channel information. In addition, a self-ensemble module is leveraged to further boost the model performance. Evaluation experimental results on the 2019 and 2020 Brain Tumor Segmentation (BraTS) datasets show that our SAResU-Net is superior to its baseline, especially on the tumor core segmentation task. Moreover, our model achieves DSC values of 79.17%, 90.02% and 82.00% for the enhancing tumor (ET), the whole tumor (WT), and tumor core(TC) on the BraTS 2020 validation dataset, respectively, while on the validation dataset of BraTS 2019, the values are 77.74%, 90.40% and 83.58%, respectively, proving its effectiveness in the application of brain tumor segmentation.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"53 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":"122668929","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-BMEI56279.2022.9980043
Jiapeng Shen, Shengxi Diao
In this paper, a bandgap voltage reference (BGR) with low temperature coefficient (TC) and high power supply rejection (PSR) is proposed. To obtain a low TC, an exponential compensation circuit is inserted to the BGR, which calibrate the high-order temperature coefficient of the base-emitter voltage $V_{BE}$ in bipolar transistor. A PSR enhancement stage is inserted to suppress supply noise. In the post-layout simulation, TC is 0.2 ppm/°C over −55°C to 125 °C and the PSR is −71.4dB@100KHz and −70.2dB@1MHz. Fabricated in 0.18um BiCMOS technology, the proposed bandgap voltage reference obtain an active area of 125um x 83um.
提出了一种具有低温系数(TC)和高电源抑制(PSR)的带隙基准电压(BGR)。为了获得低TC，在BGR中插入指数补偿电路，对双极晶体管基极-发射极电压的高阶温度系数$V_{BE}$进行校准。插入PSR增强级以抑制电源噪声。在布局后仿真中，在- 55°C至125°C范围内，TC为0.2 ppm/°C, PSR为- 71.4dB@100KHz和- 70.2dB@1MHz。采用0.18um BiCMOS技术制造，所提出的带隙电压基准的有效面积为125um x 83um。
{"title":"A Sub-1 ppm/°C TC Bandgap Voltage Reference with High Power Supply Rejection","authors":"Jiapeng Shen, Shengxi Diao","doi":"10.1109/CISP-BMEI56279.2022.9980043","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9980043","url":null,"abstract":"In this paper, a bandgap voltage reference (BGR) with low temperature coefficient (TC) and high power supply rejection (PSR) is proposed. To obtain a low TC, an exponential compensation circuit is inserted to the BGR, which calibrate the high-order temperature coefficient of the base-emitter voltage $V_{BE}$ in bipolar transistor. A PSR enhancement stage is inserted to suppress supply noise. In the post-layout simulation, TC is 0.2 ppm/°C over −55°C to 125 °C and the PSR is −71.4dB@100KHz and −70.2dB@1MHz. Fabricated in 0.18um BiCMOS technology, the proposed bandgap voltage reference obtain an active area of 125um x 83um.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"190 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":"122825770","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-BMEI56279.2022.9980005
Yan Li, Qingyuan Bai, Xianjun Yang, Xu Zhou, Yining Sun, Zhiming Yao
Gait disturbance is one of the main clinical symptoms of Parkinson's disease, mainly manifested by disrupted gait rhythm and increased variability, prone to bradykinetic, festinating, freezing, and other gaits in different stages of the disease. Wearable devices are widely used tools to monitor gait disorders in Parkinson's disease. This study aims to develop a wearable device-based system for monitoring and quantitatively analyzing multiple abnormal gait patterns, which includes wearable devices, a mobile application, and a server. The wearable device is a combination of a force-sensitive insole and an inertial measurement unit. The mobile application connects to the sensors via Bluetooth to collect the signals and transmits them to the server, which calculates the features and uses a pre-trained machine learning classifier to detect abnormalities in the patient's gait. During model training, a subset of features with 70.01 % importance of all features was retained, and the performance of three machine learning classifiers was compared for normal gait, bradykinetic gait, festinating gait, and freezing of gait, with the best results of 0.9722 recall, 0.9788 precision, 97.31 % accuracy, and 0.9755 F1-score.
{"title":"An abnormal gait monitoring system for patients with Parkinson's disease based on wearable devices","authors":"Yan Li, Qingyuan Bai, Xianjun Yang, Xu Zhou, Yining Sun, Zhiming Yao","doi":"10.1109/CISP-BMEI56279.2022.9980005","DOIUrl":"https://doi.org/10.1109/CISP-BMEI56279.2022.9980005","url":null,"abstract":"Gait disturbance is one of the main clinical symptoms of Parkinson's disease, mainly manifested by disrupted gait rhythm and increased variability, prone to bradykinetic, festinating, freezing, and other gaits in different stages of the disease. Wearable devices are widely used tools to monitor gait disorders in Parkinson's disease. This study aims to develop a wearable device-based system for monitoring and quantitatively analyzing multiple abnormal gait patterns, which includes wearable devices, a mobile application, and a server. The wearable device is a combination of a force-sensitive insole and an inertial measurement unit. The mobile application connects to the sensors via Bluetooth to collect the signals and transmits them to the server, which calculates the features and uses a pre-trained machine learning classifier to detect abnormalities in the patient's gait. During model training, a subset of features with 70.01 % importance of all features was retained, and the performance of three machine learning classifiers was compared for normal gait, bradykinetic gait, festinating gait, and freezing of gait, with the best results of 0.9722 recall, 0.9788 precision, 97.31 % accuracy, and 0.9755 F1-score.","PeriodicalId":198522,"journal":{"name":"2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)","volume":"53 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":"121824709","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：