Pub Date : 2021-10-25DOI: 10.1109/BIBE52308.2021.9635306
D. Cvetković, A. Cvetkovic, Danijela D. Nikodijević, Jovana V. Jovankić, Milena G. Milutinović, V. Stojić, N. Zdravković, Slobodanka Mltrović
Despite the obvious progress in the field of diagnosis and therapy, further measures are needed to increase the effectiveness of treatment and reduce morbidity and mortality from breast cancer. An immunofluorescence method was used to determine the protein expression of VEGF and CXCR-4 in tumor and peritumoral tissue. Peritumoral tissue is not only a passive factor, but actively participates in the process of tumor growth and development, as well as in the processes of recurrence and metastasis. Markers of neoangiogenesis in tumor and peritumoral tissue such as protein expression of VEGF and CXCR-4 receptors may serve as reliable predictors of disease outcome in breast cancer patients, which may provide useful suggestions in treatment choices.
{"title":"Determination of VEGF and CXCR4 in Tumor and Peritumoral Tissue of Patients with Breast Cancer as a Predictive Factor","authors":"D. Cvetković, A. Cvetkovic, Danijela D. Nikodijević, Jovana V. Jovankić, Milena G. Milutinović, V. Stojić, N. Zdravković, Slobodanka Mltrović","doi":"10.1109/BIBE52308.2021.9635306","DOIUrl":"https://doi.org/10.1109/BIBE52308.2021.9635306","url":null,"abstract":"Despite the obvious progress in the field of diagnosis and therapy, further measures are needed to increase the effectiveness of treatment and reduce morbidity and mortality from breast cancer. An immunofluorescence method was used to determine the protein expression of VEGF and CXCR-4 in tumor and peritumoral tissue. Peritumoral tissue is not only a passive factor, but actively participates in the process of tumor growth and development, as well as in the processes of recurrence and metastasis. Markers of neoangiogenesis in tumor and peritumoral tissue such as protein expression of VEGF and CXCR-4 receptors may serve as reliable predictors of disease outcome in breast cancer patients, which may provide useful suggestions in treatment choices.","PeriodicalId":343724,"journal":{"name":"2021 IEEE 21st International Conference on Bioinformatics and Bioengineering (BIBE)","volume":"52 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130331047","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 : 2021-10-25DOI: 10.1109/BIBE52308.2021.9635460
Charles A. Ellis, Rongen Zhang, V. Calhoun, Darwin A. Carbajal, Robyn Miller, May D. Wang
In recent years, more biomedical studies have begun to use multimodal data to improve model performance. Many studies have used ablation for explainability, which requires the modification of input data. This can create out-of-distribution samples and lead to incorrect explanations. To avoid this problem, we propose using a gradient-based feature attribution approach, called layer-wise relevance propagation (LRP), to explain the importance of modalities both locally and globally for the first time. We demonstrate the feasibility of the approach with sleep stage classification as our use-case and train a 1-D convolutional neural network with electroencephalogram (EEG), electrooculogram (EOG), and electromyogram (EMG) data. We also analyze the relationship of our local explainability results with clinical and demographic variables to determine whether they affect our classifier. Across all samples, EEG is the most important modality, followed by EOG and EMG. For individual sleep stages, EEG and EOG have higher relevance for awake and non-rapid eye movement 1 (NREM1). EOG is most important for REM, and EEG is most relevant for NREM2-NREM3. Also, LRP gives consistent levels of importance to each modality for the correctly classified samples across folds but inconsistent levels of importance for incorrectly classified samples. Our statistical analyses suggest that medication has a significant effect upon patterns learned for EEG and EOG NREM2 and that subject sex and age significantly affects the EEG and EOG patterns learned, respectively. Our results demonstrate the viability of gradient-based approaches for explaining multimodal electrophysiology classifiers and suggest their generalizability for other multimodal classification domains.
{"title":"A Gradient-based Approach for Explaining Multimodal Deep Learning Classifiers","authors":"Charles A. Ellis, Rongen Zhang, V. Calhoun, Darwin A. Carbajal, Robyn Miller, May D. Wang","doi":"10.1109/BIBE52308.2021.9635460","DOIUrl":"https://doi.org/10.1109/BIBE52308.2021.9635460","url":null,"abstract":"In recent years, more biomedical studies have begun to use multimodal data to improve model performance. Many studies have used ablation for explainability, which requires the modification of input data. This can create out-of-distribution samples and lead to incorrect explanations. To avoid this problem, we propose using a gradient-based feature attribution approach, called layer-wise relevance propagation (LRP), to explain the importance of modalities both locally and globally for the first time. We demonstrate the feasibility of the approach with sleep stage classification as our use-case and train a 1-D convolutional neural network with electroencephalogram (EEG), electrooculogram (EOG), and electromyogram (EMG) data. We also analyze the relationship of our local explainability results with clinical and demographic variables to determine whether they affect our classifier. Across all samples, EEG is the most important modality, followed by EOG and EMG. For individual sleep stages, EEG and EOG have higher relevance for awake and non-rapid eye movement 1 (NREM1). EOG is most important for REM, and EEG is most relevant for NREM2-NREM3. Also, LRP gives consistent levels of importance to each modality for the correctly classified samples across folds but inconsistent levels of importance for incorrectly classified samples. Our statistical analyses suggest that medication has a significant effect upon patterns learned for EEG and EOG NREM2 and that subject sex and age significantly affects the EEG and EOG patterns learned, respectively. Our results demonstrate the viability of gradient-based approaches for explaining multimodal electrophysiology classifiers and suggest their generalizability for other multimodal classification domains.","PeriodicalId":343724,"journal":{"name":"2021 IEEE 21st International Conference on Bioinformatics and Bioengineering (BIBE)","volume":"222 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122884896","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 : 2021-10-25DOI: 10.1109/BIBE52308.2021.9635481
G. Apostolopoulos, A. Koutras, D. Anyfantis, Ioanna Christoyianni
Computer-aided Diagnosis (CAD) systems have become a significant assistance tool, that are used to help identify abnormal/normal regions of interest in mammograms faster and more effectively than human readers. In this work, we propose a new approach for breast cancer identification of all type of lesions in digital mammograms by combining low-and high-level mammogram descriptors in a compact form. The proposed method consists of two major stages: Initially, a feature extraction process that utilizes two dimensional discrete transforms based on ART, Shapelets and textural representations based on Gabor filter banks, is used to extract low-level visual descriptors. To further improve our method's performance, the semantic information of each mammogram given by radiologists is encoded in a 16-bit length word high-level feature vector. All features are stored in a quaternion and fused using the L2 norm prior to their presentation to the classification module. For the classification task, each ROS is recognized using two different classification models, Ada Boost and Random Forest. The proposed method is evaluated on regions taken from the DDSM database. The results show that Ada Boost outperforms Random Forest in terms of accuracy (99.2%$(pm 0.527)$ against 93.78% $(pm 1.659))$, precision, recall and F-measure. Both classifiers achieve a mean accuracy of 33% and 38% higher than using only visual descriptors, showing that semantic information can indeed improve the diagnosis when it is combined with standard visual features.
{"title":"A Comparative Analysis of Breast Cancer Diagnosis by Fusing Visual and Semantic Feature Descriptors","authors":"G. Apostolopoulos, A. Koutras, D. Anyfantis, Ioanna Christoyianni","doi":"10.1109/BIBE52308.2021.9635481","DOIUrl":"https://doi.org/10.1109/BIBE52308.2021.9635481","url":null,"abstract":"Computer-aided Diagnosis (CAD) systems have become a significant assistance tool, that are used to help identify abnormal/normal regions of interest in mammograms faster and more effectively than human readers. In this work, we propose a new approach for breast cancer identification of all type of lesions in digital mammograms by combining low-and high-level mammogram descriptors in a compact form. The proposed method consists of two major stages: Initially, a feature extraction process that utilizes two dimensional discrete transforms based on ART, Shapelets and textural representations based on Gabor filter banks, is used to extract low-level visual descriptors. To further improve our method's performance, the semantic information of each mammogram given by radiologists is encoded in a 16-bit length word high-level feature vector. All features are stored in a quaternion and fused using the L2 norm prior to their presentation to the classification module. For the classification task, each ROS is recognized using two different classification models, Ada Boost and Random Forest. The proposed method is evaluated on regions taken from the DDSM database. The results show that Ada Boost outperforms Random Forest in terms of accuracy (99.2%$(pm 0.527)$ against 93.78% $(pm 1.659))$, precision, recall and F-measure. Both classifiers achieve a mean accuracy of 33% and 38% higher than using only visual descriptors, showing that semantic information can indeed improve the diagnosis when it is combined with standard visual features.","PeriodicalId":343724,"journal":{"name":"2021 IEEE 21st International Conference on Bioinformatics and Bioengineering (BIBE)","volume":"100 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134315547","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 : 2021-10-25DOI: 10.1109/BIBE52308.2021.9635230
Filippos Bitsas, Irini Georgia Dimitriou, G. Manis
Now days, new methods, ideas and applications are reinforcing safety in our home environment. Children's safety is a major concern for all parents, especially the new ones. Potential dangers are hidden everywhere, even in the children's room. Motivated by the necessity for additional safety, we employed smart technology to develop a sensor based system for reducing hazards from electricity, such as electric shocks. A smart system for additional protection was designed, targeting the periods in which parents are absent and the children alone in their room. The proposed system adds value in existing safety measures, since it works complementary to them. The main idea is based on the detection of the presence of adults in the room. Depending on parents' presence, the smart system decides which sockets are allowed to be active and which are not. Android software forwards observations on the activity to the parent's mobile phone and allows easier management. A prototype of the system has been developed and tested, without the participation of children in the experiments.
{"title":"Smart Protection from Electricity Hazards in Children's Room","authors":"Filippos Bitsas, Irini Georgia Dimitriou, G. Manis","doi":"10.1109/BIBE52308.2021.9635230","DOIUrl":"https://doi.org/10.1109/BIBE52308.2021.9635230","url":null,"abstract":"Now days, new methods, ideas and applications are reinforcing safety in our home environment. Children's safety is a major concern for all parents, especially the new ones. Potential dangers are hidden everywhere, even in the children's room. Motivated by the necessity for additional safety, we employed smart technology to develop a sensor based system for reducing hazards from electricity, such as electric shocks. A smart system for additional protection was designed, targeting the periods in which parents are absent and the children alone in their room. The proposed system adds value in existing safety measures, since it works complementary to them. The main idea is based on the detection of the presence of adults in the room. Depending on parents' presence, the smart system decides which sockets are allowed to be active and which are not. Android software forwards observations on the activity to the parent's mobile phone and allows easier management. A prototype of the system has been developed and tested, without the participation of children in the experiments.","PeriodicalId":343724,"journal":{"name":"2021 IEEE 21st International Conference on Bioinformatics and Bioengineering (BIBE)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133424800","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 : 2021-10-25DOI: 10.1109/BIBE52308.2021.9635304
Xiaozhi Yuan, Daniel Bibl, Kahlil Khan, Lei Sun
In silico approach can make vaccine designs more efficient and cost-effective. It complements the traditional process and becomes extremely valuable in coping with pandemics such as COVID-19. A recent study proposed an artificial intelligence-based framework to predict and design multi-epitope vaccines for the SARS-CoV-2 virus. However, we found several issues in its dataset design as well as its neural network design. To achieve more reliable predictions of the potential vaccine subunits, we create a more reliable and larger dataset for machine learning experiments. We apply natural language processing techniques and build neural networks composed of convolutional layer and recurrent layer to identify peptide sequences as vaccine candidates. We also train a classifier using embeddings from a pre-trained Transformer protein language model, which provides a baseline for comparison. Experimental results demonstrate that our models achieve high performance in classification accuracy and the area under the receiver operating characteristic curve.
{"title":"Predicting Multi-Epitope Vaccine Candidates Using Natural Language Processing and Deep Learning","authors":"Xiaozhi Yuan, Daniel Bibl, Kahlil Khan, Lei Sun","doi":"10.1109/BIBE52308.2021.9635304","DOIUrl":"https://doi.org/10.1109/BIBE52308.2021.9635304","url":null,"abstract":"In silico approach can make vaccine designs more efficient and cost-effective. It complements the traditional process and becomes extremely valuable in coping with pandemics such as COVID-19. A recent study proposed an artificial intelligence-based framework to predict and design multi-epitope vaccines for the SARS-CoV-2 virus. However, we found several issues in its dataset design as well as its neural network design. To achieve more reliable predictions of the potential vaccine subunits, we create a more reliable and larger dataset for machine learning experiments. We apply natural language processing techniques and build neural networks composed of convolutional layer and recurrent layer to identify peptide sequences as vaccine candidates. We also train a classifier using embeddings from a pre-trained Transformer protein language model, which provides a baseline for comparison. Experimental results demonstrate that our models achieve high performance in classification accuracy and the area under the receiver operating characteristic curve.","PeriodicalId":343724,"journal":{"name":"2021 IEEE 21st International Conference on Bioinformatics and Bioengineering (BIBE)","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133263479","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 : 2021-10-25DOI: 10.1109/BIBE52308.2021.9635424
M. Neghina, R. Petruse, S. Ćuković, Caliri Schiau, Nenad Filipović
Assessment of the spinal disorders is a notoriously difficult problem, even in controlled environments where the patients are instructed to stand upright. The method presented here considers the analysis of the mathematical curvature of the scaled and interpolated spinal line, in both the sagittal and frontal planes. Although the number of assumptions for spine normality is kept to a (reasonable) minimum, we demonstrate good detection of sharp or otherwise unnatural local bending in adolescent spinal alignments.
{"title":"Automatic Curvature Analysis for Finely Interpolated Spinal Curves","authors":"M. Neghina, R. Petruse, S. Ćuković, Caliri Schiau, Nenad Filipović","doi":"10.1109/BIBE52308.2021.9635424","DOIUrl":"https://doi.org/10.1109/BIBE52308.2021.9635424","url":null,"abstract":"Assessment of the spinal disorders is a notoriously difficult problem, even in controlled environments where the patients are instructed to stand upright. The method presented here considers the analysis of the mathematical curvature of the scaled and interpolated spinal line, in both the sagittal and frontal planes. Although the number of assumptions for spine normality is kept to a (reasonable) minimum, we demonstrate good detection of sharp or otherwise unnatural local bending in adolescent spinal alignments.","PeriodicalId":343724,"journal":{"name":"2021 IEEE 21st International Conference on Bioinformatics and Bioengineering (BIBE)","volume":"94 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122468297","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 : 2021-10-25DOI: 10.1109/BIBE52308.2021.9635269
Smiljana Tomasevic, I. Šaveljić, L. Velicki, N. Filipovic
The aim of this work was to evaluate the impact of Bicuspid Aortic Valve (BAV), on displacements, Von Mises stress, shear stress and pressure distribution within the aortic root by using computational Finite Element (FE) method. The three-dimensional (3D) patient-specific geometry of dilated aortic root with BAV was reconstructed based on Computed Tomography (CT) scan images, in order to obtain the 3D finite element mesh. Two types of analyses: i) structural analysis and ii) computational fluid dynamics (CFD) were performed, with applied equivalent material characteristics of BAV and boundary conditions. The initial results for this single case, displacements and Von Mises stress distribution (for structural analysis), as well as shear stress and pressure distribution (for CFD analysis) were quantified concerning anatomical patient's structures. The regions of abnormal stresses on the aortic leaflets and annulus, with asymmetrically open bicuspid valve, were related to the increased pressures and shear stresses and analyzed for this patient-specific case. Due to the difficulties in obtaining such characteristics in vitro or in vivo, the performed computational analysis gave better insight into the biomechanics of the aortic root with BAV that is needed to achieve improvements in surgical repair techniques and presurgical planning.
{"title":"Computational Finite Element Analysis of Aortic Root with Bicuspid Valve","authors":"Smiljana Tomasevic, I. Šaveljić, L. Velicki, N. Filipovic","doi":"10.1109/BIBE52308.2021.9635269","DOIUrl":"https://doi.org/10.1109/BIBE52308.2021.9635269","url":null,"abstract":"The aim of this work was to evaluate the impact of Bicuspid Aortic Valve (BAV), on displacements, Von Mises stress, shear stress and pressure distribution within the aortic root by using computational Finite Element (FE) method. The three-dimensional (3D) patient-specific geometry of dilated aortic root with BAV was reconstructed based on Computed Tomography (CT) scan images, in order to obtain the 3D finite element mesh. Two types of analyses: i) structural analysis and ii) computational fluid dynamics (CFD) were performed, with applied equivalent material characteristics of BAV and boundary conditions. The initial results for this single case, displacements and Von Mises stress distribution (for structural analysis), as well as shear stress and pressure distribution (for CFD analysis) were quantified concerning anatomical patient's structures. The regions of abnormal stresses on the aortic leaflets and annulus, with asymmetrically open bicuspid valve, were related to the increased pressures and shear stresses and analyzed for this patient-specific case. Due to the difficulties in obtaining such characteristics in vitro or in vivo, the performed computational analysis gave better insight into the biomechanics of the aortic root with BAV that is needed to achieve improvements in surgical repair techniques and presurgical planning.","PeriodicalId":343724,"journal":{"name":"2021 IEEE 21st International Conference on Bioinformatics and Bioengineering (BIBE)","volume":"95 3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116495222","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 : 2021-10-25DOI: 10.1109/BIBE52308.2021.9635163
Aishwarya Purohit, S. Acharya, James Green
Computational prediction of protein-protein interaction (PPI) from protein sequence is important as many cellular functions are made possible through PPI. The Protein Interaction Prediction Engine (PIPE) software suite was developed for such predictions. The specific location of interaction is predicted by the PIPE-Sites predictor, which depends on PIPE engine. This PIPE-Sites predictor is here updated through the use of a large high-quality dataset of known PPI sites. Additionally, a similarity-weighted score had been recently developed in PIPE4 and has been proven to be more accurate for the likelihood of PPI prediction. However, PIPE-Sites are shown to be ineffective when applied to similarity-weighted score data. Thus, we here propose and evaluate a new sequence-based PPI site prediction method, named Panorama. This new method leverages similarity-weighted score data to further increase performance over two different performance metrics when evaluated on both $boldsymbol{H}$. sapiens and $boldsymbol{S}$, cerevisiae PPI site data.
{"title":"A novel Greedy approach for Sequence based Computational prediction of Binding-Sites in Protein-Protein Interaction","authors":"Aishwarya Purohit, S. Acharya, James Green","doi":"10.1109/BIBE52308.2021.9635163","DOIUrl":"https://doi.org/10.1109/BIBE52308.2021.9635163","url":null,"abstract":"Computational prediction of protein-protein interaction (PPI) from protein sequence is important as many cellular functions are made possible through PPI. The Protein Interaction Prediction Engine (PIPE) software suite was developed for such predictions. The specific location of interaction is predicted by the PIPE-Sites predictor, which depends on PIPE engine. This PIPE-Sites predictor is here updated through the use of a large high-quality dataset of known PPI sites. Additionally, a similarity-weighted score had been recently developed in PIPE4 and has been proven to be more accurate for the likelihood of PPI prediction. However, PIPE-Sites are shown to be ineffective when applied to similarity-weighted score data. Thus, we here propose and evaluate a new sequence-based PPI site prediction method, named Panorama. This new method leverages similarity-weighted score data to further increase performance over two different performance metrics when evaluated on both $boldsymbol{H}$. sapiens and $boldsymbol{S}$, cerevisiae PPI site data.","PeriodicalId":343724,"journal":{"name":"2021 IEEE 21st International Conference on Bioinformatics and Bioengineering (BIBE)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124546686","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 : 2021-10-25DOI: 10.1109/BIBE52308.2021.9635351
Theofilos Zacheilas, K. Moirogiorgou, N. Papandroulakis, E. Sotiriades, M. Zervakis, A. Dollas
Aquaculture faces the issue of net integrity on cage farming. Holes on the net need to be detected but as yet the process is not fully automated. This work is a second-generation embedded system to detect in real time holes in aquaculture nets from a video input. It extends previous results by processing video rather than still images, under lighting variation, haze, and different size of holes along each frame. The modeling and simulation of the new algorithm has been done in MATLAB; the system has been designed and implemented on a Field Programmable Gate Array (FPGA) - based platform. The proposed system has substantially better performance vs. software at a much lower energy consumption.
{"title":"An FPGA-Based System for Video Processing to Detect Holes in Aquaculture Nets","authors":"Theofilos Zacheilas, K. Moirogiorgou, N. Papandroulakis, E. Sotiriades, M. Zervakis, A. Dollas","doi":"10.1109/BIBE52308.2021.9635351","DOIUrl":"https://doi.org/10.1109/BIBE52308.2021.9635351","url":null,"abstract":"Aquaculture faces the issue of net integrity on cage farming. Holes on the net need to be detected but as yet the process is not fully automated. This work is a second-generation embedded system to detect in real time holes in aquaculture nets from a video input. It extends previous results by processing video rather than still images, under lighting variation, haze, and different size of holes along each frame. The modeling and simulation of the new algorithm has been done in MATLAB; the system has been designed and implemented on a Field Programmable Gate Array (FPGA) - based platform. The proposed system has substantially better performance vs. software at a much lower energy consumption.","PeriodicalId":343724,"journal":{"name":"2021 IEEE 21st International Conference on Bioinformatics and Bioengineering (BIBE)","volume":" 45","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120828903","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 : 2021-10-25DOI: 10.1109/BIBE52308.2021.9635506
A. Vukicevic, A. Zabotti, V. Milic, A. Hočevar, O. Lucia, G. Filippou, A. Tzioufas, S. Vita, Nenad Filipović
Salivary gland ultrasonography (SGUS) represents a promising tool for diagnosing Primary Sjögren's syndrome (pSS), which is manifest with abnormalities in salivary glands (SG). In this study, we propose a fully automatic method for scoring SGs in SGUS images, which is the most important step towards SG the pSS diagnosis. A two-centric cohort included 600 images (150 patients) annotated by experienced clinicians. The aim of the study was to assess various deep learning classifiers (MobileNetV2, VGG19, Dense-Net, Squeeze-Net, Inception_v3, and ResNet) for the purpose of the pSS scoring in SGUS. The training was performed using the ADAM optimizer and cross entropy loss function. Top performing algorithms were MobileNetV2, ResNet, and Dense-Net. The assessment showed that deep learning algorithms reached clinicians-level performances in the almost real-time. Considering that, the further work should be regarded towards evaluation on larger and international data sets with the goal to establish SGUS as an effective noninvasive pSS diagnostic tool.
{"title":"Scoring Primary Sjögren's syndrome affected salivary glands ultrasonography images by using deep learning algorithms","authors":"A. Vukicevic, A. Zabotti, V. Milic, A. Hočevar, O. Lucia, G. Filippou, A. Tzioufas, S. Vita, Nenad Filipović","doi":"10.1109/BIBE52308.2021.9635506","DOIUrl":"https://doi.org/10.1109/BIBE52308.2021.9635506","url":null,"abstract":"Salivary gland ultrasonography (SGUS) represents a promising tool for diagnosing Primary Sjögren's syndrome (pSS), which is manifest with abnormalities in salivary glands (SG). In this study, we propose a fully automatic method for scoring SGs in SGUS images, which is the most important step towards SG the pSS diagnosis. A two-centric cohort included 600 images (150 patients) annotated by experienced clinicians. The aim of the study was to assess various deep learning classifiers (MobileNetV2, VGG19, Dense-Net, Squeeze-Net, Inception_v3, and ResNet) for the purpose of the pSS scoring in SGUS. The training was performed using the ADAM optimizer and cross entropy loss function. Top performing algorithms were MobileNetV2, ResNet, and Dense-Net. The assessment showed that deep learning algorithms reached clinicians-level performances in the almost real-time. Considering that, the further work should be regarded towards evaluation on larger and international data sets with the goal to establish SGUS as an effective noninvasive pSS diagnostic tool.","PeriodicalId":343724,"journal":{"name":"2021 IEEE 21st International Conference on Bioinformatics and Bioengineering (BIBE)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-10-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130371675","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：