Pub Date : 2021-09-13DOI: 10.23919/cinc53138.2021.9662851
M. Taconné, V. Rolle, K. Owashi, V. Panis, A. Hubert, V. Auffret, E. Galli, Alfredo I. Hernández, E. Donal
The objective of this study is to propose a model-based method, adapted to patients with severe aortic stenosis (AS), in order to reproduce left ventricle (LV) pressure and volume from patient specific data. A formal sensitivity analysis is proposed, focused on left ventricle volume and pressure. The most influent parameters of this analysis are then selected to be identified in a parameter identification strategy and provide a patient specific pressure curve. This was implemented on 3 AS patients and a close match was observed between experimental and simulated pressure and volume curves. The global root mean square error (RMSE) for pressure and volume curves are respectively 21.8 $(pm 1.8)$ mmHg and 14.8 $(pm 9.4)ml$,. The model-based approach proposed shows promising results to generate accurate LV pressure and volume in AS case.
{"title":"Sensitivity Analysis and Parameter Identification of a Cardiovascular Model in Aortic Stenosis","authors":"M. Taconné, V. Rolle, K. Owashi, V. Panis, A. Hubert, V. Auffret, E. Galli, Alfredo I. Hernández, E. Donal","doi":"10.23919/cinc53138.2021.9662851","DOIUrl":"https://doi.org/10.23919/cinc53138.2021.9662851","url":null,"abstract":"The objective of this study is to propose a model-based method, adapted to patients with severe aortic stenosis (AS), in order to reproduce left ventricle (LV) pressure and volume from patient specific data. A formal sensitivity analysis is proposed, focused on left ventricle volume and pressure. The most influent parameters of this analysis are then selected to be identified in a parameter identification strategy and provide a patient specific pressure curve. This was implemented on 3 AS patients and a close match was observed between experimental and simulated pressure and volume curves. The global root mean square error (RMSE) for pressure and volume curves are respectively 21.8 $(pm 1.8)$ mmHg and 14.8 $(pm 9.4)ml$,. The model-based approach proposed shows promising results to generate accurate LV pressure and volume in AS case.","PeriodicalId":126746,"journal":{"name":"2021 Computing in Cardiology (CinC)","volume":"60 1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114489947","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 : 2021-09-13DOI: 10.23919/cinc53138.2021.9662714
J. V. Prehn, Svetoslav Ivanov, G. Nalbantov
Automated detection of key cardiac pathologies in reduced-lead ECGs is an enabling factor in applying ECG analysis on a larger scale. The PhysioNet/Computing in Cardiology Challenge 2021 identifies a set of key cardiac pathologies and challenges us with the task to automatically detect them. Critical to this task is the extraction of features from these ECGs which, combined, mark the presence of one or more of these key cardiac pathologies. Methodology: algorithms were devised to automatically extract features based on the definitions as used in medical practice, beat morphology and image deformation. A binary classifier for each key cardiac pathology was trained using these features, extracted from the labeled ECGs from The Challenge. The binary classifiers were combined into a multi-label classifier by learning thresholds on the scores of the binary classifiers using Bayesian optimization in a cross-validation setting. Results: our contribution submitted for evaluation achieved a challenge metric score of 0.28, 0.31, 0.32, 0.28 and 0.23 placing us (team DSC) 29, 25, 25, 28 and 28 out of 39 teams which submitted an official entry on 12-, 6-,4-, 3- and 2-lead test datasets respectively.
在低导联心电图中自动检测关键心脏病理是在更大范围内应用ECG分析的一个有利因素。PhysioNet/Computing in Cardiology Challenge 2021确定了一组关键的心脏病理，并挑战我们自动检测它们的任务。这项任务的关键是从这些心电图中提取特征，这些特征结合起来，标志着一种或多种关键心脏病理的存在。方法:设计算法，根据医学实践中使用的定义，beat形态学和图像变形自动提取特征。使用这些特征训练每个关键心脏病理的二元分类器，这些特征从the Challenge的标记心电图中提取。通过在交叉验证设置中使用贝叶斯优化学习二元分类器分数的阈值，将二元分类器组合成多标签分类器。结果:我们提交评估的贡献达到了0.28、0.31、0.32、0.28和0.23的挑战度量得分，在提交12、6、4、3和2领先测试数据集的39个正式参赛团队中，我们(DSC团队)分别排名29、25、25、28和28。
{"title":"Pathologies Prediction on Short ECG Signals with Focus on Feature Extraction Based on Beat Morphology and Image Deformation","authors":"J. V. Prehn, Svetoslav Ivanov, G. Nalbantov","doi":"10.23919/cinc53138.2021.9662714","DOIUrl":"https://doi.org/10.23919/cinc53138.2021.9662714","url":null,"abstract":"Automated detection of key cardiac pathologies in reduced-lead ECGs is an enabling factor in applying ECG analysis on a larger scale. The PhysioNet/Computing in Cardiology Challenge 2021 identifies a set of key cardiac pathologies and challenges us with the task to automatically detect them. Critical to this task is the extraction of features from these ECGs which, combined, mark the presence of one or more of these key cardiac pathologies. Methodology: algorithms were devised to automatically extract features based on the definitions as used in medical practice, beat morphology and image deformation. A binary classifier for each key cardiac pathology was trained using these features, extracted from the labeled ECGs from The Challenge. The binary classifiers were combined into a multi-label classifier by learning thresholds on the scores of the binary classifiers using Bayesian optimization in a cross-validation setting. Results: our contribution submitted for evaluation achieved a challenge metric score of 0.28, 0.31, 0.32, 0.28 and 0.23 placing us (team DSC) 29, 25, 25, 28 and 28 out of 39 teams which submitted an official entry on 12-, 6-,4-, 3- and 2-lead test datasets respectively.","PeriodicalId":126746,"journal":{"name":"2021 Computing in Cardiology (CinC)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125725262","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 : 2021-09-13DOI: 10.23919/cinc53138.2021.9662704
Wenjie Cai, Fanli Liu, Xuan Wang, Bo-Ming Xu, Yao-Chin Wang
Introduction: The electrocardiogram (ECG) is the most common diagnostic tool for screening cardiovascular diseases. PhysioNet/Computing in Cardiology Challenge 2021 aims to classify cardiac abnormalities from twelve-lead, six-lead, four-lead, three-lead, and two-lead ECGs. Methods: ECGs were downsampled to 250 Hz and then applied with a bandpass filter to reduce noise. The unscored label named VEB was transformed to PVC. The ECGs labeled as AF in the Ningbo Database were relabeled as AFL or AF. All ECGs were randomly shuffled and divided into a training set and a validation set at 4:1. Five models based on a deep residual convolutional neural network were proposed to make classification from different dimensions of ECGs. A novel loss calculation method was proposed to balance the different labeling tendency of different source data sets. Results: Our team, USST_Med, received an official test score of 0.54, 0.52, 0.50, 0.51, and 0.50 on twelve-lead, six-lead, four-lead, three-lead, and two-lead ECG test sets, respectively. The scores are ranked 5th, 3rd, 7th, 5th and 7th, respectively. Conclusion: The proposed models performed well on classifying ECGs and have potential for clinical application.
心电图(ECG)是筛查心血管疾病最常用的诊断工具。PhysioNet/Computing in Cardiology Challenge 2021旨在对12导联、6导联、4导联、3导联和2导联心电图中的心脏异常进行分类。方法:将心电图降采样至250 Hz，然后应用带通滤波器降低噪声。命名为VEB的未得分标签被转化为PVC。将宁波数据库中标记为AF的心电图重新标记为AFL或AF。所有心电图随机洗牌，按4:1划分为训练集和验证集。提出了基于深度残差卷积神经网络的5种模型，对不同维度的脑电图进行分类。为了平衡不同源数据集标注倾向的差异，提出了一种新的损失计算方法。结果:我们的团队USST_Med在12导联、6导联、4导联、3导联和2导联心电图测试仪上的官方测试分数分别为0.54、0.52、0.50、0.51和0.50。排名分别为第5、第3、第7、第5、第7。结论:所建立的模型对脑电图具有较好的分类效果，具有临床应用价值。
{"title":"Classifying Different Dimensional ECGs Using Deep Residual Convolutional Neural Networks","authors":"Wenjie Cai, Fanli Liu, Xuan Wang, Bo-Ming Xu, Yao-Chin Wang","doi":"10.23919/cinc53138.2021.9662704","DOIUrl":"https://doi.org/10.23919/cinc53138.2021.9662704","url":null,"abstract":"Introduction: The electrocardiogram (ECG) is the most common diagnostic tool for screening cardiovascular diseases. PhysioNet/Computing in Cardiology Challenge 2021 aims to classify cardiac abnormalities from twelve-lead, six-lead, four-lead, three-lead, and two-lead ECGs. Methods: ECGs were downsampled to 250 Hz and then applied with a bandpass filter to reduce noise. The unscored label named VEB was transformed to PVC. The ECGs labeled as AF in the Ningbo Database were relabeled as AFL or AF. All ECGs were randomly shuffled and divided into a training set and a validation set at 4:1. Five models based on a deep residual convolutional neural network were proposed to make classification from different dimensions of ECGs. A novel loss calculation method was proposed to balance the different labeling tendency of different source data sets. Results: Our team, USST_Med, received an official test score of 0.54, 0.52, 0.50, 0.51, and 0.50 on twelve-lead, six-lead, four-lead, three-lead, and two-lead ECG test sets, respectively. The scores are ranked 5th, 3rd, 7th, 5th and 7th, respectively. Conclusion: The proposed models performed well on classifying ECGs and have potential for clinical application.","PeriodicalId":126746,"journal":{"name":"2021 Computing in Cardiology (CinC)","volume":"2006 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125835936","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 : 2021-09-13DOI: 10.23919/cinc53138.2021.9662739
Hansheng Ren, Miao Xiong, Bryan Hooi
In PhysioNet/Computing in Cardiology Challenge 2021, we developed an ensemble model by combining different epochs of ResNet to classify cardiac abnormalities from 12,6,4,3,2 lead electrocardiogram (ECG) signals, where epochs are chosen based on validation performance on China Physiological Signal Challenge (CPSC) dataset and Georgia dataset. In order to adapt to the specially designed Challenge score, we designed a multi-task loss to combine the benefit of binary cross-entropy loss and Challenge loss. Besides, we also integrated a subsample frequency feature into the model to learn from the signals. To gain a better generalization ability, mixup and weighted loss are introduced. We submitted our model in the official phase with team name DataLA_NUS, and our final selected model achieved a Challenge score of 0.51, 0.51, 0.51, 0.50, 0.52 (ranked 8th, 5th, 6th, 8th, 5th) on the 12-lead, 6-lead, 4-lead, 3-lead, and 2-lead setting on the final hidden test set with the Challenge evaluation metric.
在PhysioNet/Computing In Cardiology Challenge 2021中，我们通过结合ResNet的不同时间点开发了一个集成模型，对12、6、4、3、2导联心电图(ECG)信号进行心脏异常分类，其中时间点的选择是基于中国生理信号挑战(CPSC)数据集和Georgia数据集的验证性能。为了适应专门设计的挑战分数，我们设计了一种多任务损失，将二值交叉熵损失和挑战损失的优点结合起来。此外，我们还将子样本频率特征集成到模型中以从信号中学习。为了获得更好的泛化能力，引入了混合和加权损失。我们在正式阶段以团队名称DataLA_NUS提交了我们的模型，最终选择的模型在最终隐藏测试集的12领先、6领先、4领先、3领先和2领先设置下获得了0.51、0.51、0.51、0.51、0.50、0.52(排名第8、5、6、8、5)的Challenge分数。
{"title":"Robust and Task-Aware Training of Deep Residual Networks for Varying-Lead ECG Classification","authors":"Hansheng Ren, Miao Xiong, Bryan Hooi","doi":"10.23919/cinc53138.2021.9662739","DOIUrl":"https://doi.org/10.23919/cinc53138.2021.9662739","url":null,"abstract":"In PhysioNet/Computing in Cardiology Challenge 2021, we developed an ensemble model by combining different epochs of ResNet to classify cardiac abnormalities from 12,6,4,3,2 lead electrocardiogram (ECG) signals, where epochs are chosen based on validation performance on China Physiological Signal Challenge (CPSC) dataset and Georgia dataset. In order to adapt to the specially designed Challenge score, we designed a multi-task loss to combine the benefit of binary cross-entropy loss and Challenge loss. Besides, we also integrated a subsample frequency feature into the model to learn from the signals. To gain a better generalization ability, mixup and weighted loss are introduced. We submitted our model in the official phase with team name DataLA_NUS, and our final selected model achieved a Challenge score of 0.51, 0.51, 0.51, 0.50, 0.52 (ranked 8th, 5th, 6th, 8th, 5th) on the 12-lead, 6-lead, 4-lead, 3-lead, and 2-lead setting on the final hidden test set with the Challenge evaluation metric.","PeriodicalId":126746,"journal":{"name":"2021 Computing in Cardiology (CinC)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126417187","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 : 2021-09-13DOI: 10.23919/cinc53138.2021.9662885
Haoyu Jiang, Mimi Hu, Junbiao Hong, Yijing Li, Xianliang He
In this paper, we analysed the features of pacing pulses and challenging noises from clinical datasets collected at high sampling rate. A two-stage algorithm is proposed to detect pacing pulses for real-time application purpose. In the first stage, pulse candidates were picked up preliminarily after enhancing the rising and falling edges of the pulses and attenuating high frequency noises. More detailed morphology features were checked in the second stage to validate and confirm the candidates. The sensitivity and positive predictivity of the algorithm on the training and testing datasets both exceed 99%. The evaluation results illustrate the pretty good performance of the proposed algorithm.
{"title":"A Real-Time Digital Pacemaker Pulse Detection Algorithm","authors":"Haoyu Jiang, Mimi Hu, Junbiao Hong, Yijing Li, Xianliang He","doi":"10.23919/cinc53138.2021.9662885","DOIUrl":"https://doi.org/10.23919/cinc53138.2021.9662885","url":null,"abstract":"In this paper, we analysed the features of pacing pulses and challenging noises from clinical datasets collected at high sampling rate. A two-stage algorithm is proposed to detect pacing pulses for real-time application purpose. In the first stage, pulse candidates were picked up preliminarily after enhancing the rising and falling edges of the pulses and attenuating high frequency noises. More detailed morphology features were checked in the second stage to validate and confirm the candidates. The sensitivity and positive predictivity of the algorithm on the training and testing datasets both exceed 99%. The evaluation results illustrate the pretty good performance of the proposed algorithm.","PeriodicalId":126746,"journal":{"name":"2021 Computing in Cardiology (CinC)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123715178","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 : 2021-09-13DOI: 10.23919/cinc53138.2021.9662862
Jenny Venton
Deep learning models for electrocardiogram (ECG) classification can be affected by the presence of physiological noise on the ECG, as shown in previous work. In this study, we explore the impact of different physiological noise types, and differing signal-to-noise ratios (SNRs) of noise on classification performance. We find that classification performance is impacted differently by different noise types. In addition, the best classification performance comes from using a network trained on clean ECGs to classify clean ECGs. In conclusion, this study has revealed several questions regarding inclusion or exclusion of noise on the ECG for training and classification by deep learning models.
{"title":"Investigating the Robustness of Deep Learning to Electrocardiogram Noise","authors":"Jenny Venton","doi":"10.23919/cinc53138.2021.9662862","DOIUrl":"https://doi.org/10.23919/cinc53138.2021.9662862","url":null,"abstract":"Deep learning models for electrocardiogram (ECG) classification can be affected by the presence of physiological noise on the ECG, as shown in previous work. In this study, we explore the impact of different physiological noise types, and differing signal-to-noise ratios (SNRs) of noise on classification performance. We find that classification performance is impacted differently by different noise types. In addition, the best classification performance comes from using a network trained on clean ECGs to classify clean ECGs. In conclusion, this study has revealed several questions regarding inclusion or exclusion of noise on the ECG for training and classification by deep learning models.","PeriodicalId":126746,"journal":{"name":"2021 Computing in Cardiology (CinC)","volume":"106 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121946545","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 : 2021-09-13DOI: 10.23919/cinc53138.2021.9662748
Daniel Gedon, Antônio H. Ribeiro, Niklas Wahlström, Thomas Bo Schön
Self-supervised learning is a paradigm that extracts general features which describe the input space by artificially generating labels from the input without the need for explicit annotations. The learned features can then be used by transfer learning to boost the performance on a downstream task. Such methods have recently produced state of the art results in natural language processing and computer vision. Here, we propose a self-supervised learning method for 12-lead electrocardiograms (ECGs). For pretraining the model we design a task to mask out subsegements of all channels of the input signals and try to predict the actual values. As the model architecture, we use a U-ResNet containing an encoder-decoder structure. We test our method by self-supervised pretraining on the CODE dataset and then transfer the learnt features by finetuning on the PTB-XL and CPSC benchmarks to evaluate the effect of our method in the classification of 12-leads ECGs. The method does provide modest improvements in performance when compared to not using pretraining. In future work we will make use of these ideas in smaller dataset, where we believe it can lead to larger performance gains.
{"title":"First Steps Towards Self-Supervised Pretraining of the 12-Lead ECG","authors":"Daniel Gedon, Antônio H. Ribeiro, Niklas Wahlström, Thomas Bo Schön","doi":"10.23919/cinc53138.2021.9662748","DOIUrl":"https://doi.org/10.23919/cinc53138.2021.9662748","url":null,"abstract":"Self-supervised learning is a paradigm that extracts general features which describe the input space by artificially generating labels from the input without the need for explicit annotations. The learned features can then be used by transfer learning to boost the performance on a downstream task. Such methods have recently produced state of the art results in natural language processing and computer vision. Here, we propose a self-supervised learning method for 12-lead electrocardiograms (ECGs). For pretraining the model we design a task to mask out subsegements of all channels of the input signals and try to predict the actual values. As the model architecture, we use a U-ResNet containing an encoder-decoder structure. We test our method by self-supervised pretraining on the CODE dataset and then transfer the learnt features by finetuning on the PTB-XL and CPSC benchmarks to evaluate the effect of our method in the classification of 12-leads ECGs. The method does provide modest improvements in performance when compared to not using pretraining. In future work we will make use of these ideas in smaller dataset, where we believe it can lead to larger performance gains.","PeriodicalId":126746,"journal":{"name":"2021 Computing in Cardiology (CinC)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121698647","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 : 2021-09-13DOI: 10.23919/cinc53138.2021.9662874
Jackie H Boynton, Byung Suk Lee
Once a cardiac alarm is triggered in the intensive care unit (ICU), accurately classifying whether the alarm is true of false is of critical importance. Incorrect classification may lead to patient's death if the alarm is true or to disruption in patient care if false. There has been a body of research, as signified by the 2015 PhysioNet/CinC Challenge; due accomplishments have been made in the relevant computational technology, and yet the highest accuracy known thus far is in the mid-80% range (85%). Our work achieved much higher accuracy and, additionally, very early classification almost at the onset of an arrhythmia alarm, by utilizing state of the art deep learning methods. The machine learning model used is a Residual Network (ResNet) and a Bi-directional Long Short Term Memory (BiLSTM) connected in tandem. Using the Phy-sioNet dataset of 750 recorded ECG segments published with the Challenge, our method performed the classification with 96% accuracy in 0.52 seconds from the onset of an alarm on average over all test ECG segments.
{"title":"Deep Learning Based Classification of True/False Arrhythmia Alarms in the Intensive Care Unit","authors":"Jackie H Boynton, Byung Suk Lee","doi":"10.23919/cinc53138.2021.9662874","DOIUrl":"https://doi.org/10.23919/cinc53138.2021.9662874","url":null,"abstract":"Once a cardiac alarm is triggered in the intensive care unit (ICU), accurately classifying whether the alarm is true of false is of critical importance. Incorrect classification may lead to patient's death if the alarm is true or to disruption in patient care if false. There has been a body of research, as signified by the 2015 PhysioNet/CinC Challenge; due accomplishments have been made in the relevant computational technology, and yet the highest accuracy known thus far is in the mid-80% range (85%). Our work achieved much higher accuracy and, additionally, very early classification almost at the onset of an arrhythmia alarm, by utilizing state of the art deep learning methods. The machine learning model used is a Residual Network (ResNet) and a Bi-directional Long Short Term Memory (BiLSTM) connected in tandem. Using the Phy-sioNet dataset of 750 recorded ECG segments published with the Challenge, our method performed the classification with 96% accuracy in 0.52 seconds from the onset of an alarm on average over all test ECG segments.","PeriodicalId":126746,"journal":{"name":"2021 Computing in Cardiology (CinC)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123403174","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 : 2021-09-13DOI: 10.23919/cinc53138.2021.9662692
Victor Gonçalves Marques, A. Gharaviri, Simone Pezzuto, P. Bonizzi, S. Zeemering, U. Schotten
Catheter ablation treatment for atrial fibrillation (AF) is still suboptimal, possibly due to the difficulty to identify AF drivers. Recurrence analysis can be used to detect and eventually locate repetitive patterns that tend to be generated by AF drivers. In this study, we aimed to understand the spatial relationship between repetitiveness in recurrence analysis and rotor positions in an in-silico AF model. AF was simulated in a detailed three-dimensional model of the atria considering different degrees of endomysial fibrosis (0% and 70%). Rotors driving AF were tracked based on phase singularities obtained from transmembrane potentials. Activation-phase signals calculated from electrograms (4×4 electrode grid, 3 mm spacing) were used for recurrence analysis. Intervals with and without long-lasting sources inside the electrode coverage area were determined; the recurrence in both groups of intervals was quantified and compared with each other by calculating the recurrence rate (RR) per AF cycle length. RRs were lower during intervals with sources for both 0% and 70% fibrosis groups (0.56 [0.36;0.85] vs. 0.90 [0.80;0.97], $p < 0.001$ and 0.73 [0.41;0.84] vs. 0.87 [0.76;0.92], $p < 0.001$, respectively). These results indicate that recurrences are found in the area adjacent to the sources but not on the sources themselves, thus suggesting that recurrence analysis could contribute to guide ablation therapy.
{"title":"Spatial Relationship Between Atrial Fibrillation Drivers and the Presence of Repetitive Conduction Patterns Using Recurrence Analysis on In-Silico Models","authors":"Victor Gonçalves Marques, A. Gharaviri, Simone Pezzuto, P. Bonizzi, S. Zeemering, U. Schotten","doi":"10.23919/cinc53138.2021.9662692","DOIUrl":"https://doi.org/10.23919/cinc53138.2021.9662692","url":null,"abstract":"Catheter ablation treatment for atrial fibrillation (AF) is still suboptimal, possibly due to the difficulty to identify AF drivers. Recurrence analysis can be used to detect and eventually locate repetitive patterns that tend to be generated by AF drivers. In this study, we aimed to understand the spatial relationship between repetitiveness in recurrence analysis and rotor positions in an in-silico AF model. AF was simulated in a detailed three-dimensional model of the atria considering different degrees of endomysial fibrosis (0% and 70%). Rotors driving AF were tracked based on phase singularities obtained from transmembrane potentials. Activation-phase signals calculated from electrograms (4×4 electrode grid, 3 mm spacing) were used for recurrence analysis. Intervals with and without long-lasting sources inside the electrode coverage area were determined; the recurrence in both groups of intervals was quantified and compared with each other by calculating the recurrence rate (RR) per AF cycle length. RRs were lower during intervals with sources for both 0% and 70% fibrosis groups (0.56 [0.36;0.85] vs. 0.90 [0.80;0.97], $p < 0.001$ and 0.73 [0.41;0.84] vs. 0.87 [0.76;0.92], $p < 0.001$, respectively). These results indicate that recurrences are found in the area adjacent to the sources but not on the sources themselves, thus suggesting that recurrence analysis could contribute to guide ablation therapy.","PeriodicalId":126746,"journal":{"name":"2021 Computing in Cardiology (CinC)","volume":"74 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126473429","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 : 2021-09-13DOI: 10.23919/cinc53138.2021.9662957
Muhammad Usman Gul, K. Kadir, Muhammad Haziq Kamarul Azman
In the previous study, the atrial flutter mechanism (i.e., Focal or Macroreentrant) was differentiated from the standard 12-lead ECG by the variability of the cycle length of visible successive P-waves (between the R-R waves). This study aims to help researchers reduce imbalances through two different techniques, especially in atrial flutter. Besides, early detection of the AFL mechanism can increase the efficacy of invasive elimination. The proposed model has been extracted several features derived from statistical analysis of the intervals of successive atrial rhythm. Forty-eight patients were undergone endoscopic catheter ablation for the identifications of the AFL mechanism. Two different techniques, SMOTE and Smoothed-Bootstrap, have been used to augment and re-balance the dataset. The synthetic data generated by Smoothed-Bootstrap has been much closer to the original dataset and relatively better than SMOTE technique. The performance has been evaluated by three linear classifiers Linear Discriminant Analysis (LDA), Logistic Regression (LOG), and Support Vector Machine (SVM). The LOG classifier achieved its average performance with accuracy, specificity, sensitivity, 71.08%, 77.13%, and 65.12%, respectively. Smoothed-Bootstrap is a suitable technique in AFL cases to minimize the imbalance issue. The variability in cycle length of consecutive P-waves from the surface ECG has differentiated the Focal AFLfrom Macrorrentrant AFL.
{"title":"Data Augmentation for Discrimination of Atrial Flutter Mechanism Using 12-Lead Surface Electrocardiogram","authors":"Muhammad Usman Gul, K. Kadir, Muhammad Haziq Kamarul Azman","doi":"10.23919/cinc53138.2021.9662957","DOIUrl":"https://doi.org/10.23919/cinc53138.2021.9662957","url":null,"abstract":"In the previous study, the atrial flutter mechanism (i.e., Focal or Macroreentrant) was differentiated from the standard 12-lead ECG by the variability of the cycle length of visible successive P-waves (between the R-R waves). This study aims to help researchers reduce imbalances through two different techniques, especially in atrial flutter. Besides, early detection of the AFL mechanism can increase the efficacy of invasive elimination. The proposed model has been extracted several features derived from statistical analysis of the intervals of successive atrial rhythm. Forty-eight patients were undergone endoscopic catheter ablation for the identifications of the AFL mechanism. Two different techniques, SMOTE and Smoothed-Bootstrap, have been used to augment and re-balance the dataset. The synthetic data generated by Smoothed-Bootstrap has been much closer to the original dataset and relatively better than SMOTE technique. The performance has been evaluated by three linear classifiers Linear Discriminant Analysis (LDA), Logistic Regression (LOG), and Support Vector Machine (SVM). The LOG classifier achieved its average performance with accuracy, specificity, sensitivity, 71.08%, 77.13%, and 65.12%, respectively. Smoothed-Bootstrap is a suitable technique in AFL cases to minimize the imbalance issue. The variability in cycle length of consecutive P-waves from the surface ECG has differentiated the Focal AFLfrom Macrorrentrant AFL.","PeriodicalId":126746,"journal":{"name":"2021 Computing in Cardiology (CinC)","volume":"219 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121469397","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：