Fenqi Rong, Banghua Yang, Jun Ma, Shouwei Gao, Xinxing Xia
Brain-computer interfaces (BCI) enable direct communication with external equipment, using neural activity as the control signal. Electroencephalogram (EEG) signals are usually selected as the control signal. For EEG signals obtained from a given experimental paradigm, a superior algorithm for feature extraction and classification is very significant. As one of the representative algorithms of deep learning, the convolutional neural network (CNN) has been widely used in the field of BCI. In this work, we introduce the filter-bank convolutional network (FBCNet) and propose an improved method. It mainly improves the network performance by modifying the loss function. The single loss function in the network is improved to the multi-loss fusion functions. Various loss functions are added to the network, and the characteristics of different loss functions are used to train the network to improve the network classification performance. This method of multi-loss fusion functions is validated on a dataset of 11 healthy subjects and compared with the other three benchmark algorithms. The result shows that the improved FBCNet produces a four-classes accuracy of 78.5%, which is superior to other algorithms.
{"title":"Motor imagery EEG decoding based on multi-loss fusion FBCNet","authors":"Fenqi Rong, Banghua Yang, Jun Ma, Shouwei Gao, Xinxing Xia","doi":"10.1145/3563737.3563755","DOIUrl":"https://doi.org/10.1145/3563737.3563755","url":null,"abstract":"Brain-computer interfaces (BCI) enable direct communication with external equipment, using neural activity as the control signal. Electroencephalogram (EEG) signals are usually selected as the control signal. For EEG signals obtained from a given experimental paradigm, a superior algorithm for feature extraction and classification is very significant. As one of the representative algorithms of deep learning, the convolutional neural network (CNN) has been widely used in the field of BCI. In this work, we introduce the filter-bank convolutional network (FBCNet) and propose an improved method. It mainly improves the network performance by modifying the loss function. The single loss function in the network is improved to the multi-loss fusion functions. Various loss functions are added to the network, and the characteristics of different loss functions are used to train the network to improve the network classification performance. This method of multi-loss fusion functions is validated on a dataset of 11 healthy subjects and compared with the other three benchmark algorithms. The result shows that the improved FBCNet produces a four-classes accuracy of 78.5%, which is superior to other algorithms.","PeriodicalId":127021,"journal":{"name":"Proceedings of the 7th International Conference on Biomedical Signal and Image Processing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130019970","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}
In recent years with the advent of big data, unsupervised feature extraction has developed rapidly, among which independent component analysis (ICA), as a classical unsupervised technique, has been widely applied in a variety of data scenarios. This paper proposes an incremental unsupervised feature extraction method based on one specific kind of ICA, i.e. Infomax. Specifically, an incremental singular value decomposition (SVD) was used in combination with the a hierarchical Infomax principle, so as to realize the rapid batch processing of data and reduce the computational complexity. Then, this method was tested with MNIST, a handwritten data set for experimental verification. The results showed that the proposed method can greatly improve the speed of feature extraction under the condition of large data volume, and ensure that the calculation results are consistent with the previous training method. Furthermore, by application in Google Speech Recognition Challenge, we verified that this method can significantly improve the training efficiency for real-world pattern recognition scenarios. The proposed method can be applied in feature extraction, data visualization and supervised learning of high-dimensional data.
{"title":"An incremental unsupervised feature extraction method based on Infomax","authors":"Weikun Niu, Sen Yuan, Feng Zhang","doi":"10.1145/3563737.3563747","DOIUrl":"https://doi.org/10.1145/3563737.3563747","url":null,"abstract":"In recent years with the advent of big data, unsupervised feature extraction has developed rapidly, among which independent component analysis (ICA), as a classical unsupervised technique, has been widely applied in a variety of data scenarios. This paper proposes an incremental unsupervised feature extraction method based on one specific kind of ICA, i.e. Infomax. Specifically, an incremental singular value decomposition (SVD) was used in combination with the a hierarchical Infomax principle, so as to realize the rapid batch processing of data and reduce the computational complexity. Then, this method was tested with MNIST, a handwritten data set for experimental verification. The results showed that the proposed method can greatly improve the speed of feature extraction under the condition of large data volume, and ensure that the calculation results are consistent with the previous training method. Furthermore, by application in Google Speech Recognition Challenge, we verified that this method can significantly improve the training efficiency for real-world pattern recognition scenarios. The proposed method can be applied in feature extraction, data visualization and supervised learning of high-dimensional data.","PeriodicalId":127021,"journal":{"name":"Proceedings of the 7th International Conference on Biomedical Signal and Image Processing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132089774","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}
The oral clefts are abnormality that occurs to fetus during the development of lip and palate, which have affected numerous infants. The major problems associated with cleft lips and palates involve causing poor oral hygiene and increasing the chances of cavities. In order to prevent fetus from such an innate disease, diagnosis during pregnancy is extremely necessary. Moreover, the diagnosis of oral clefts must be completed during pregnancy and without major invasion to the pregnant, in order to avoid potential harm to babies. Therefore, ultrasound imaging, which is radiation-free, low-cost, and harmless, has been the predominant diagnosing modality since its invention. However, the resolving capability of ultrasound imaging is affected by many factors, such as noises and artifacts, which contribute to false diagnosis during medical examination, resulting in biased and potentially fake results. This strong limitation has caused a significant amount of babies that are failed to be confirmed. Therefore, understanding the limitation of ultrasound imaging is extremely important in terms of diagnosing oral clefts. Quantifying the limitation with living subjects is often problematic as individuals differ from each other and cause undesired inconsistency. To overcome this challenge, medical phantom is usually designed to evaluate and analyze different imaging modalities. Here, we designed an ultrasound imaging phantom using 3-dimentional printing, and then quantified the limiting sonography resolution by imaging it in aqueous environment. We concluded that there are certain limitations of different resolutions, which better illustrates the false diagnosis of oral clefts and palates. These results provide a reference criteria for using sonography to diagnose oral clefts.
{"title":"Determination of Resolution Limitation of Sonography used in Diagnosis of Cleft Lips and Palates","authors":"Yuyao Jiang, Virgia Wang","doi":"10.1145/3563737.3563742","DOIUrl":"https://doi.org/10.1145/3563737.3563742","url":null,"abstract":"The oral clefts are abnormality that occurs to fetus during the development of lip and palate, which have affected numerous infants. The major problems associated with cleft lips and palates involve causing poor oral hygiene and increasing the chances of cavities. In order to prevent fetus from such an innate disease, diagnosis during pregnancy is extremely necessary. Moreover, the diagnosis of oral clefts must be completed during pregnancy and without major invasion to the pregnant, in order to avoid potential harm to babies. Therefore, ultrasound imaging, which is radiation-free, low-cost, and harmless, has been the predominant diagnosing modality since its invention. However, the resolving capability of ultrasound imaging is affected by many factors, such as noises and artifacts, which contribute to false diagnosis during medical examination, resulting in biased and potentially fake results. This strong limitation has caused a significant amount of babies that are failed to be confirmed. Therefore, understanding the limitation of ultrasound imaging is extremely important in terms of diagnosing oral clefts. Quantifying the limitation with living subjects is often problematic as individuals differ from each other and cause undesired inconsistency. To overcome this challenge, medical phantom is usually designed to evaluate and analyze different imaging modalities. Here, we designed an ultrasound imaging phantom using 3-dimentional printing, and then quantified the limiting sonography resolution by imaging it in aqueous environment. We concluded that there are certain limitations of different resolutions, which better illustrates the false diagnosis of oral clefts and palates. These results provide a reference criteria for using sonography to diagnose oral clefts.","PeriodicalId":127021,"journal":{"name":"Proceedings of the 7th International Conference on Biomedical Signal and Image Processing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124105782","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}
The detection of carotid artery in ultrasound medical images helps locating other organs to build an efficient computer-aided diagnosis system. Due to the differences in the speed, direction and angle of continuous scanning of the carotid artery, its imaging shape in the ultrasound image is complex and easy to be distorted. Meanwhile, the labeled medical image datasets are limited. So it's hard to make carotid artery detection accurately. Although existing methods such as Mask R-CNN attempt to achieve carotid artery segmentation by ultrasound data, the results are not ideal. We propose a novel architecture called SSL-DH-Faster RCNN, which is based on a semi-supervised learning approach using unlabeled medical images to improve our model performance. In our framework, we adopt double head detection architecture to solve the problem that single head structure performs poorly on handling both classification and localization task at the same time. Concretely, a fully connected head(fc-head) for classification task and a convolution head(conv-head) for regression is adopted based on the reason that fc-head got better performance on classification task and conv-head is more suitable for localization task. Simultaneously, we combine PAFPN module into our framework to make low-layer information easier to propagate with above two methods and improve model performance further. Experiments show that SSL-DH-Faster RCNN method proposed in this paper achieves superior performance, and outperforms several popular methods. Experiments show that compared with existing popular methods, our method achieves the best performance on AP50, AP75 and AP@[0.50:0.95] metrics.
{"title":"Semi-supervised learning with double head approach for carotid artery detection","authors":"Zhiwei Li, Wei Peng, Changquan Lu","doi":"10.1145/3563737.3563740","DOIUrl":"https://doi.org/10.1145/3563737.3563740","url":null,"abstract":"The detection of carotid artery in ultrasound medical images helps locating other organs to build an efficient computer-aided diagnosis system. Due to the differences in the speed, direction and angle of continuous scanning of the carotid artery, its imaging shape in the ultrasound image is complex and easy to be distorted. Meanwhile, the labeled medical image datasets are limited. So it's hard to make carotid artery detection accurately. Although existing methods such as Mask R-CNN attempt to achieve carotid artery segmentation by ultrasound data, the results are not ideal. We propose a novel architecture called SSL-DH-Faster RCNN, which is based on a semi-supervised learning approach using unlabeled medical images to improve our model performance. In our framework, we adopt double head detection architecture to solve the problem that single head structure performs poorly on handling both classification and localization task at the same time. Concretely, a fully connected head(fc-head) for classification task and a convolution head(conv-head) for regression is adopted based on the reason that fc-head got better performance on classification task and conv-head is more suitable for localization task. Simultaneously, we combine PAFPN module into our framework to make low-layer information easier to propagate with above two methods and improve model performance further. Experiments show that SSL-DH-Faster RCNN method proposed in this paper achieves superior performance, and outperforms several popular methods. Experiments show that compared with existing popular methods, our method achieves the best performance on AP50, AP75 and AP@[0.50:0.95] metrics.","PeriodicalId":127021,"journal":{"name":"Proceedings of the 7th International Conference on Biomedical Signal and Image Processing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124745957","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}
Attention deficit hyperactivity disorder (ADHD) is a common neurodevelopmental disorder in teenagers. Some excellent ADHD automatic diagnosis system extracted features from magnetic resonance image (MRI). Researchers have shown fMRI data offers specific measure of ADHD brain activity. In this paper, we propose a low-frequency fluctuation feature map generation approach for ADHD diagnosis, which can highlight the discriminative parts of fMRI features. However, the extracted feature maps still have redundant information. So we add the attention mechanism which can pay more attention to the local information. In order to successfully apply the attention mechanism to convolutional neural network (CNN) and match it to 3D fMRI feature maps, we extend convolutional block attention module (CBAM) from 2D plane to 3D geometric space. After that, we design a single modality 3D CNN based on 3D CBAM to diagnosis ADHD via low-frequency fluctuation feature map. Our model is evaluated on ADHD-200 dataset and it obtains the state-of-the-art classification accuracy of 75.83%. At the same time, our model also simplifies the feature extraction module and the classification module of multi-modality method.
{"title":"ADHD Classification With Low-Frequency Fluctuation Feature Map Based on 3D CBAMe","authors":"Lihua Su, Sei-ichiro Kamata","doi":"10.1145/3563737.3563749","DOIUrl":"https://doi.org/10.1145/3563737.3563749","url":null,"abstract":"Attention deficit hyperactivity disorder (ADHD) is a common neurodevelopmental disorder in teenagers. Some excellent ADHD automatic diagnosis system extracted features from magnetic resonance image (MRI). Researchers have shown fMRI data offers specific measure of ADHD brain activity. In this paper, we propose a low-frequency fluctuation feature map generation approach for ADHD diagnosis, which can highlight the discriminative parts of fMRI features. However, the extracted feature maps still have redundant information. So we add the attention mechanism which can pay more attention to the local information. In order to successfully apply the attention mechanism to convolutional neural network (CNN) and match it to 3D fMRI feature maps, we extend convolutional block attention module (CBAM) from 2D plane to 3D geometric space. After that, we design a single modality 3D CNN based on 3D CBAM to diagnosis ADHD via low-frequency fluctuation feature map. Our model is evaluated on ADHD-200 dataset and it obtains the state-of-the-art classification accuracy of 75.83%. At the same time, our model also simplifies the feature extraction module and the classification module of multi-modality method.","PeriodicalId":127021,"journal":{"name":"Proceedings of the 7th International Conference on Biomedical Signal and Image Processing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124485649","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}
This paper proposed a new ECG denoising approach based on singular value decomposition (SVD) for EMG (electromyogram) artifacts reduction. Unlike the traditional method like discrete wavelet transform and classical bandpass filter weakly noise reduction performance by frequency domain overlapping, we propose to remove EMG by an improved adaptive matrix construction and achieve high signal to noise ratio (SNR), the ECG signal were firstly spitted a number of heartbeats to construct the trajectory matrix; then the trajectory matrix was decomposed by SVD; at last, the decision rules used to reconstruct the clear signal, the proposed method is evaluated on the MIT-BIH arrhythmia database, the result shows our method attain a high improvement of signal to noise ratio output (SNR) and lower signal distortion.
{"title":"An Improved Adaptive Periodical Segment Matrix for Processing EMG Artifacts in ECG Signal Detection","authors":"Xing Liu, Zhiming Long, Zhuqing Wang","doi":"10.1145/3563737.3563757","DOIUrl":"https://doi.org/10.1145/3563737.3563757","url":null,"abstract":"This paper proposed a new ECG denoising approach based on singular value decomposition (SVD) for EMG (electromyogram) artifacts reduction. Unlike the traditional method like discrete wavelet transform and classical bandpass filter weakly noise reduction performance by frequency domain overlapping, we propose to remove EMG by an improved adaptive matrix construction and achieve high signal to noise ratio (SNR), the ECG signal were firstly spitted a number of heartbeats to construct the trajectory matrix; then the trajectory matrix was decomposed by SVD; at last, the decision rules used to reconstruct the clear signal, the proposed method is evaluated on the MIT-BIH arrhythmia database, the result shows our method attain a high improvement of signal to noise ratio output (SNR) and lower signal distortion.","PeriodicalId":127021,"journal":{"name":"Proceedings of the 7th International Conference on Biomedical Signal and Image Processing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132393277","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}
Abstract: Objective: To analyze the molecular mechanism of Taohong Siwu Decoction in the treatment of diabetic heart failure based on network pharmacology and bioinformatics technology.METHODS: The bioactive components of Taohong Siwu Decoction were screened by TCMSP, a database of traditional Chinese medicine systems pharmacology, and the targets of the active components were predicted by Swiss Target Prediction. At the same time, the GEO database was searched for data sets related to diabetic heart failure, the data set GSE26887 was used for research, and the GEO2R online analysis tool and R language were used for differential gene screening and annotation. The drug targets and disease targets were imported into Cytoscape to construct a protein-protein interaction (PPI) network to obtain key genes. The key genes were imported into the Metascape platform for GO enrichment analysis and KEGG signaling pathway analysis. Results: A total of 49 active ingredients of Taohong Siwu Decoction and 754 potential therapeutic targets were obtained. Differential gene screening was performed on the dataset GSE26887, and 69 significantly expressed genes were obtained. 754 drug targets and 69 disease targets were imported into Cytoscape for protein-protein interaction, and BisoGenet plug-in was used for topological parameter analysis, and 323 key targets of Taohong Siwu Decoction in the treatment of diabetic heart failure were obtained. Conclusion: Taohong Siwu Decoction in the treatment of diabetic heart failure has the characteristics of multiple components, multiple pathways and multiple targets. Among them, the key genes are NTRK1, HSP90AA1, CUL3, TUBA4A, TP53. Important pathways are estrogen signaling pathway, ErbB signaling pathway, p53 signaling pathway, and Hedgehog signaling pathway. They may play a combined role in the treatment of diabetic heart failure.
{"title":"To explore the mechanism of Taohong Siwu Decoction on diabetic heart failure based on GEO differential gene chip data and network pharmacology","authors":"K. Cao, Wei Wang, Junli Zhang, Lei Deng, F. Han","doi":"10.1145/3563737.3563745","DOIUrl":"https://doi.org/10.1145/3563737.3563745","url":null,"abstract":"Abstract: Objective: To analyze the molecular mechanism of Taohong Siwu Decoction in the treatment of diabetic heart failure based on network pharmacology and bioinformatics technology.METHODS: The bioactive components of Taohong Siwu Decoction were screened by TCMSP, a database of traditional Chinese medicine systems pharmacology, and the targets of the active components were predicted by Swiss Target Prediction. At the same time, the GEO database was searched for data sets related to diabetic heart failure, the data set GSE26887 was used for research, and the GEO2R online analysis tool and R language were used for differential gene screening and annotation. The drug targets and disease targets were imported into Cytoscape to construct a protein-protein interaction (PPI) network to obtain key genes. The key genes were imported into the Metascape platform for GO enrichment analysis and KEGG signaling pathway analysis. Results: A total of 49 active ingredients of Taohong Siwu Decoction and 754 potential therapeutic targets were obtained. Differential gene screening was performed on the dataset GSE26887, and 69 significantly expressed genes were obtained. 754 drug targets and 69 disease targets were imported into Cytoscape for protein-protein interaction, and BisoGenet plug-in was used for topological parameter analysis, and 323 key targets of Taohong Siwu Decoction in the treatment of diabetic heart failure were obtained. Conclusion: Taohong Siwu Decoction in the treatment of diabetic heart failure has the characteristics of multiple components, multiple pathways and multiple targets. Among them, the key genes are NTRK1, HSP90AA1, CUL3, TUBA4A, TP53. Important pathways are estrogen signaling pathway, ErbB signaling pathway, p53 signaling pathway, and Hedgehog signaling pathway. They may play a combined role in the treatment of diabetic heart failure.","PeriodicalId":127021,"journal":{"name":"Proceedings of the 7th International Conference on Biomedical Signal and Image Processing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132795397","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}
Uterine cancer is the 4th most common cancer among women, with about 66,570 new cases in the United States every year. Late-stage uterine cancer patients have a less than 20% chance of survival due to limited effectiveness in treatment options. Immunotherapy is an emerging type of cancer treatment; however, it is only effective in a subtype of uterine cancer (∼20%). Major histocompatibility complex I (MHC-I) molecules have been researched as a main mechanism assisting cancerous cells to evade death by immune cell destruction. The goal of this project is to identify molecular regulators of MHC-I in uterine cancer to aid immunotherapy. Here, we analyzed the prognostic value of MHC-I molecules based on patient and molecular datasets from The Cancer Genome Atlas. MHC-I combined with T cell markers is associated with better prognosis in uterine cancer. Two molecular candidates, interferon regulatory factor 1 (IRF1) and proteasome subunit beta type-9 (PSMB9), were identified as potential MHC-I regulators. Wet-lab experiments confirmed the role of IRF1 in regulating MHC-I expression, though PSMB9 was found to be ineffective. Furthermore, uterine cancer expressed lower levels of IRF1 compared with normal uterine tissues. This finding brings significant insight into a potential immunotherapy target molecule for treating uterine cancer. Future development includes direct testing of T cell immune responses with IRF1 enhancements to prove its effectiveness on immune cell action.
{"title":"Restoring MHC-I Molecules to Potentiate Immunotherapy in Uterine Cancer","authors":"Ashley Jiayi Zhou, Chunbo He","doi":"10.1145/3563737.3563748","DOIUrl":"https://doi.org/10.1145/3563737.3563748","url":null,"abstract":"Uterine cancer is the 4th most common cancer among women, with about 66,570 new cases in the United States every year. Late-stage uterine cancer patients have a less than 20% chance of survival due to limited effectiveness in treatment options. Immunotherapy is an emerging type of cancer treatment; however, it is only effective in a subtype of uterine cancer (∼20%). Major histocompatibility complex I (MHC-I) molecules have been researched as a main mechanism assisting cancerous cells to evade death by immune cell destruction. The goal of this project is to identify molecular regulators of MHC-I in uterine cancer to aid immunotherapy. Here, we analyzed the prognostic value of MHC-I molecules based on patient and molecular datasets from The Cancer Genome Atlas. MHC-I combined with T cell markers is associated with better prognosis in uterine cancer. Two molecular candidates, interferon regulatory factor 1 (IRF1) and proteasome subunit beta type-9 (PSMB9), were identified as potential MHC-I regulators. Wet-lab experiments confirmed the role of IRF1 in regulating MHC-I expression, though PSMB9 was found to be ineffective. Furthermore, uterine cancer expressed lower levels of IRF1 compared with normal uterine tissues. This finding brings significant insight into a potential immunotherapy target molecule for treating uterine cancer. Future development includes direct testing of T cell immune responses with IRF1 enhancements to prove its effectiveness on immune cell action.","PeriodicalId":127021,"journal":{"name":"Proceedings of the 7th International Conference on Biomedical Signal and Image Processing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128446446","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}
In previous studies, Sensory Motor Rhythm (SMR) and Movement-Related Cortical Potential (MRCP) have been proved to be complementary in decoding a variety of motion information. However, no studies have reported whether they are complementary when subjects perform functional lower limb movements. In this work, we investigate the effect of two features or their combination on classifying three functional lower limb movements (standing, walking, sitting) and rest. MRCP features are extracted by Locality Preserving Projection (LPP) and SMR features are extracted by selecting the best frequency-channel pairs through the Bhattacharyya distance. A Support Vector Machine (SVM) classifier was employed to assess the performance of different features or their combination in six binary classification tasks, where three types of lower limb movements are compared with each other or with rest. The combination of two features achieved the highest accuracy in most classification task. In the classification of standing and walking, the combination of these two features has shown significantly better performance (both p < 0.05) than the classifiers using either MRCP or SMR. Our results suggest that MRCP and SMR features are complementary for decoding the functional lower limb movements, which would benefit the Brain-computer Interface (BCI) system for lower limb rehabilitation.
{"title":"Enhance decoding of functional lower-limb movements by combining sensory motor rhythm and movement-related cortical potential features","authors":"Yulong Peng, Chenyang Li, Xuchao Chen, Xiaomeng Miao, Shaomin Zhang","doi":"10.1145/3563737.3563756","DOIUrl":"https://doi.org/10.1145/3563737.3563756","url":null,"abstract":"In previous studies, Sensory Motor Rhythm (SMR) and Movement-Related Cortical Potential (MRCP) have been proved to be complementary in decoding a variety of motion information. However, no studies have reported whether they are complementary when subjects perform functional lower limb movements. In this work, we investigate the effect of two features or their combination on classifying three functional lower limb movements (standing, walking, sitting) and rest. MRCP features are extracted by Locality Preserving Projection (LPP) and SMR features are extracted by selecting the best frequency-channel pairs through the Bhattacharyya distance. A Support Vector Machine (SVM) classifier was employed to assess the performance of different features or their combination in six binary classification tasks, where three types of lower limb movements are compared with each other or with rest. The combination of two features achieved the highest accuracy in most classification task. In the classification of standing and walking, the combination of these two features has shown significantly better performance (both p < 0.05) than the classifiers using either MRCP or SMR. Our results suggest that MRCP and SMR features are complementary for decoding the functional lower limb movements, which would benefit the Brain-computer Interface (BCI) system for lower limb rehabilitation.","PeriodicalId":127021,"journal":{"name":"Proceedings of the 7th International Conference on Biomedical Signal and Image Processing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128571969","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}
Background: Machine learning-based approaches can provide quantitative identification of the cognitive status of the brain by fMRI, which is essential to evaluate human mental activities. However, the performance of traditional machine learning algorithms is not optimal.. Methods: The data was retrieved from an open fMRI dataset of movie-watching fMRI data. Specifically, dynamic functional connectivity analysis (DFC) was calculated using a sliding-window algorithm. A gradient boosting machine learning approach was used with the DFC matrices as the features to predict the cognitive status of the human brain. Conclusion: The area under the curve (AUC) of the gradient boosting classifier with DFC measures was higher than that using conventional machine learning methods. Our findings are expected to provide a better theoretical basis for the neural mechanisms underlying cognitive status of the human brain and shed light on future machine learning-aided mental health. Risk and Safety: There are no significant risk and safety concerns in this study.
{"title":"Prediction of Cognitive Status from the Resting-State fMRI Data by Machine Learning","authors":"Qiyan Mao, Cheng Wang","doi":"10.1145/3563737.3563750","DOIUrl":"https://doi.org/10.1145/3563737.3563750","url":null,"abstract":"Background: Machine learning-based approaches can provide quantitative identification of the cognitive status of the brain by fMRI, which is essential to evaluate human mental activities. However, the performance of traditional machine learning algorithms is not optimal.. Methods: The data was retrieved from an open fMRI dataset of movie-watching fMRI data. Specifically, dynamic functional connectivity analysis (DFC) was calculated using a sliding-window algorithm. A gradient boosting machine learning approach was used with the DFC matrices as the features to predict the cognitive status of the human brain. Conclusion: The area under the curve (AUC) of the gradient boosting classifier with DFC measures was higher than that using conventional machine learning methods. Our findings are expected to provide a better theoretical basis for the neural mechanisms underlying cognitive status of the human brain and shed light on future machine learning-aided mental health. Risk and Safety: There are no significant risk and safety concerns in this study.","PeriodicalId":127021,"journal":{"name":"Proceedings of the 7th International Conference on Biomedical Signal and Image Processing","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-08-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131272203","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：