Accurate segmentation of biomedical time-series, such as intracardiac electrograms, is vital for understanding physiological states and supporting clinical interventions. Traditional rule-based and feature engineering approaches often struggle with complex clinical patterns and noise. Recent deep learning advancements offer solutions, showing various benefits and drawbacks in segmentation tasks. This study evaluates five segmentation algorithms, from traditional rule-based methods to advanced deep learning models, using a unique clinical dataset of intracardiac signals from 100 patients. We compared a rule-based method, a support vector machine (SVM), fully convolutional semantic neural network (UNet), region proposal network (Faster R-CNN), and recurrent neural network for electrocardiographic signals (DENS-ECG). Notably, Faster R-CNN has never been applied to 1D signals segmentation before. Each model underwent Bayesian optimization to minimize hyperparameter bias. Results indicated that deep learning models outperformed traditional methods, with UNet achieving the highest segmentation score of 88.9 % (root mean square errors for onset and offset of 8.43 ms and 7.49 ms), closely followed by DENS-ECG at 87.8 %. Faster R-CNN and SVM showed moderate performance, while the rule-based method had the lowest accuracy (77.7 %). UNet and DENS-ECG excelled in capturing detailed features and handling noise, highlighting their potential for clinical application. Despite greater computational demands, their superior performance and diagnostic potential support further exploration in biomedical time-series analysis.

1st Department of Internal Medicine Cardioangiology St Anne's Hospital Masaryk University Pekarska 664 53 Brno 60200 Czech Republic

Children's Hospital Faculty Hospital Brno Cernopolni 9 Brno 61300 Czech Republic

Department of Biomedical Engineering Brno University of Technology Technicka 12 Brno 61600 Czech Republic

International Clinical Research Centre St Anna's Faculty Hospital Pekarska 53 Brno 60200 Czech Republic

Zobrazit více v PubMed

Ivaturi, P. et al. A comprehensive explanation framework for biomedical time series classification. PubMed DOI PMC

Alcaine, A. et al. A wavelet-based electrogram onset delineator for automatic ventricular activation mapping. PubMed DOI

Martinez, J., Almeida, R., Olmos, S., Rocha, A. & Laguna, P. A wavelet-based ECG delineator: evaluation on standard databases. PubMed DOI

Peimankar, A. & Puthusserypady, S. DENS-ECG: A deep learning approach for ECG signal delineation. DOI

Issa, Z. F., Miller, J. M. & Zipes, D. P.

Kaiser, D. W. et al. The precise timing of tachycardia entrainment is determined by the postpacing interval, the tachycardia cycle length, and the pacing rate: Theoretical insights and practical applications. PubMed DOI PMC

Shenasa, M., Hindricks, G., Callans, D. J., Miller, J. M. & Josephson, M. E.

Pascale, P. et al. Pattern and timing of the coronary sinus activation to guide rapid diagnosis of atrial tachycardia after atrial fibrillation ablation. PubMed DOI

Steven, D., Seiler, J., Roberts-Thomson, K. C., Inada, K. & Stevenson, W. G. Mapping of atrial tachycardias after catheter ablation for atrial fibrillation: Use of bi-atrial activation patterns to facilitate recognition of origin. PubMed DOI

Haïssaguerre, M. et al. Localized Sources Maintaining Atrial Fibrillation Organized by Prior Ablation. PubMed DOI

Hongo, R. Catheter Ablation of Scar-mediated Ventricular Tachycardia: Are Substrate-based Approaches Replacing Mapping?. PubMed DOI PMC

Fustes, O. J. H. et al. Somatosensory evoked potentials in clinical practice: a review. PubMed DOI

Lantz, G. et al. Space-oriented segmentation and 3-dimensional source reconstruction of ictal EEG patterns. PubMed DOI

Costa-Garcia, A., Itkonen, M., Yamasaki, H., Shibata-Alnajjar, F. & Shimoda, S. A Novel Approach to the Segmentation of sEMG Data Based on the Activation and Deactivation of Muscle Synergies During Movement. DOI

McKinnon, M. L., Hill, N. J., Carp, J. S., Dellenbach, B. & Thompson, A. K. Methods for automated delineation and assessment of EMG responses evoked by peripheral nerve stimulation in diagnostic and closed-loop therapeutic applications. PubMed DOI PMC

Clayson, P. E., Baldwin, S. A. & Larson, M. J. ow does noise affect amplitude and latency measurement of event-related potentials (ERPs)? A methodological critique and simulation study. PubMed DOI

Paradeshi, K., Scholar, R. & Kolekar, P. D. U. Removal of EMG Artifacts from Multichannel EEG Signal Using Automatic Dynamic Segmentation and Adaptive Thresholding with Multilevel Decomposed Wavelets. DOI

Cazzaniga, G. et al. Improving the Annotation Process in Computational Pathology: A Pilot Study with Manual and Semi-automated Approaches on Consumer and Medical Grade Devices. PubMed DOI PMC

Tizhoosh, H. R. et al. Searching Images for Consensus. PubMed DOI

Khalifa, M. & Albadawy, M. AI in diagnostic imaging: Revolutionising accuracy and efficiency. DOI

Faust, O., Hagiwara, Y., Hong, T. J., Lih, O. S. & Acharya, U. R. Deep learning for healthcare applications based on physiological signals: A review. PubMed DOI

Estes, N. M. DOI

El Haddad, M. et al. Algorithmic detection of the beginning and end of bipolar electrograms: Implications for novel methods to assess local activation time during atrial tachycardia. DOI

Nothstein, M. et al. CVAR-Seg: An Automated Signal Segmentation Pipeline for Conduction Velocity and Amplitude Restitution. PubMed DOI PMC

Sánchez, J. et al. Using Machine Learning to Characterize Atrial Fibrotic Substrate From Intracardiac Signals With a Hybrid in silico and in vivo Dataset. PubMed DOI PMC

Ředina, R. et al. Arrhythmia Database with Annotated Intracardial Atrial Signals from Pediatric Patients Undergoing Catheter Ablation. 10.22489/CinC.2022.282 (2022).

Hejc, J., Redina, R., Kolarova, J. & Starek, Z. Epycon: A Single-Platform Python Package for Parsing and Converting Raw Electrophysiology Data into Open Formats. 10.22489/CinC.2023.315 (2023).

Wang, Z.-Y., Xiang, Z.-R., Zhi, J.-Y., Ding, T.-C. & Zou, R. A novel physiological signal denoising method coupled with multispectral adaptive wavelet denoising(MAWD) and unsupervised source counting algorithm(USCA). DOI

Hejč, J., Vítek, M., Ronzhina, M., Nováková, M. & Kolářová, J. A Wavelet-Based ECG Delineation Method: Adaptation to an Experimental Electrograms with Manifested Global Ischemia. PubMed DOI

Daubechies, I. The wavelet transform, time-frequency localization and signal analysis. DOI

Lee, G., Gommers, R., Waselewski, F., Wohlfahrt, K. & O’Leary, A. PyWavelets: A Python package for wavelet analysis. DOI

Ouali, M. A., Ghanai, M. & Chafaa, K. Upper envelope detection of ECG signals for baseline wander correction: a pilot study. DOI

Benitez, D., Gaydecki, P., Zaidi, A. & Fitzpatrick, A. The use of the Hilbert transform in ECG signal analysis. PubMed DOI

Burges, C. J. A Tutorial on Support Vector Machines for Pattern Recognition. DOI

Nasiri, J. A., Naghibzadeh, M., Yazdi, H. S. & Naghibzadeh, B. ECG Arrhythmia Classification with Support Vector Machines and Genetic Algorithm. In

Rabee, A. & Barhumi, I. ECG signal classification using support vector machine based on wavelet multiresolution analysis. In

Turnip, A., Ilham Rizqywan, M., Kusumandari, D. E., Turnip, M. & Sihombing, P. Classification of ECG signal with Support Vector Machine Method for Arrhythmia Detection. DOI

Smisek, R. et al. Multi-stage SVM approach for cardiac arrhythmias detection in short single-lead ECG recorded by a wearable device. PubMed DOI

LeCun, Y., Bengio, Y. & Hinton, G. Deep learning. PubMed DOI

Farabet, C., Couprie, C., Najman, L. & LeCun, Y. Learning Hierarchical Features for Scene Labeling. PubMed DOI

Krizhevsky, A., Sutskever, I. & Hinton, G. E. ImageNet classification with deep convolutional neural networks. DOI

Hinton, G. et al. Deep Neural Networks for Acoustic Modeling in Speech Recognition: The Shared Views of Four Research Groups. DOI

Leung, M. K. K., Xiong, H. Y., Lee, L. J. & Frey, B. J. Deep learning of the tissue-regulated splicing code. PubMed DOI PMC

Ma, J., Sheridan, R. P., Liaw, A., Dahl, G. E. & Svetnik, V. Deep Neural Nets as a Method for Quantitative Structure-Activity Relationships. PubMed DOI

Somani, S. et al. Deep learning and the electrocardiogram: review of the current state-of-the-art. PubMed DOI PMC

Stracina, T., Ronzhina, M., Redina, R. & Novakova, M. Golden Standard or Obsolete Method? Review of ECG Applications in Clinical and Experimental Context. PubMed DOI PMC

Hejc, J., Redina, R., Kolarova, J. & Starek, Z. Multi-channel delineation of intracardiac electrograms for arrhythmia substrate analysis using implicitly regularized convolutional neural network with wide receptive field. DOI

Ren, S., He, K., Girshick, R. & Sun, J. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks (2016). ArXiv:1506.01497 [cs]. PubMed

Ronneberger, O., Fischer, P. & Brox, T. U-Net: Convolutional Networks for Biomedical Image Segmentation, 10.48550/ARXIV.1505.04597 (2015). Version Number: 1.

Chen, L.-C., Papandreou, G., Kokkinos, I., Murphy, K. & Yuille, A. L. DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs, 10.48550/ARXIV.1606.00915 (2016). Version Number: 2. PubMed

Abraham, N. & Khan, N. M. A Novel Focal Tversky loss function with improved Attention U-Net for lesion segmentation, 10.48550/ARXIV.1810.07842 (2018). Version Number: 1.

Han, S., Eom, H., Kim, J. & Park, C. Optimal DNN architecture search using Bayesian Optimization Hyperband for arrhythmia detection. In

Akiba, T., Sano, S., Yanase, T., Ohta, T. & Koyama, M. Optuna: A Next-generation Hyperparameter Optimization Framework, 10.48550/ARXIV.1907.10902 (2019). Version Number: 1.

Kwak, S. G. & Kim, J. H. Central limit theorem: the cornerstone of modern statistics. PubMed DOI PMC

Callaert, H. & Janssen, P. The Berry-Esseen Theorem for U-Statistics. DOI

Zhou, Y., Zhu, Y. & Wong, W. K. Statistical tests for homogeneity of variance for clinical trials and recommendations. PubMed DOI PMC

Dubé, B. et al. Automatic detection and classification of human epicardial atrial unipolar electrograms. PubMed DOI

Trigano, T. & Luengo, D. Intracardiac ECG pulse localization using overlapping block sparse reconstruction. DOI

Gong, S. et al. Dilated FCN: Listening Longer to Hear Better, 10.48550/ARXIV.1907.11956 (2019). Version Number: 1.

Kim, K., Chalidabhongse, T. H., Harwood, D. & Davis, L. Real-time foreground-background segmentation using codebook model. DOI

Parks, D. H. & Fels, S. S. Evaluation of Background Subtraction Algorithms with Post-Processing. In

Waheed, K., Weaver, K. & Salam, F. A robust algorithm for detecting speech segments using an entropic contrast. In

Blanco-Velasco, M., Weng, B. & Barner, K. E. ECG signal denoising and baseline wander correction based on the empirical mode decomposition. PubMed DOI

Martínez, M., Ródenas, J., Alcaraz, R. & Rieta, J. J. Study on the Alternatives to Reduce High-Frequency Noise from Invasive Recordings of Atrial Fibrillation. 10.22489/CinC.2017.111-048 (2017).

Alwosheel, A., Van Cranenburgh, S. & Chorus, C. G. Is your dataset big enough? Sample size requirements when using artificial neural networks for discrete choice analysis. DOI

Veshchezerova, D., Bars, C. & Seitz”, J. Cycle Length Estimation Using Accurate Adaptive Detection of Local Activations in Atrial Intracardiac Electrograms. 10.22489/CinC.2022.142 (2022).

Liao, S. et al. Deep learning classification of unipolar electrograms in human atrial fibrillation: application in focal source mapping. PubMed DOI PMC

Rodrigo, M., Rogers, A. J., Ganesan, P., Alhusseini, M. & Narayan, S. M. Abstract 14742: Deep Learning of Intracardiac Electrograms in Atrial Arrhythmia.

Rodrigo, M. et al. Machine learning classifies intracardiac electrograms of atrial fibrillation from other arrhythmias. DOI