Spectral and Time Domain Parameters for The Classification of Atrial Fibrillation

Batista, Diana; Fred, Ana

doi:10.5220/0005283403290337

Cited by 2 publications

(5 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our study achieved relatively consistent accuracy values in comparison with other similar studies. The study from Batista et al [12] reported the highest accuracy of 99.08% among the studies. However, the dataset described for validation was highly unbalanced between categories.…”

Section: Svm Classifier Vs 1d-cnn Algorithmmentioning

confidence: 90%

“…The following is a literature review of Machine Learning algorithms for the task of cardiac arrhythmia feature extraction and classification. ECG classification research can be divided into either heartbeat [11][12][13] or arrhythmia [5,10,12]. The most common algorithms included in studies are Support Vector Machine (SVM) [11][12][13][14], ANN [12], and CNN [10,[14][15][16].…”

Section: Literature Reviewmentioning

confidence: 99%

“…ECG classification research can be divided into either heartbeat [11][12][13] or arrhythmia [5,10,12]. The most common algorithms included in studies are Support Vector Machine (SVM) [11][12][13][14], ANN [12], and CNN [10,[14][15][16]. Large public databases of annotated ECG signals play a fundamental role in developing algorithms for automatic ECG interpretation and classification, serving as a benchmark for comparing the validation and quantitative evaluation of algorithms from different papers in the scientific community [2].…”

Section: Literature Reviewmentioning

confidence: 99%

“…Large public databases of annotated ECG signals play a fundamental role in developing algorithms for automatic ECG interpretation and classification, serving as a benchmark for comparing the validation and quantitative evaluation of algorithms from different papers in the scientific community [2]. The frequent databases used are MIT-BIH [11,12,14,16], TNMG [10], and cinc17 [5,14,15,17].…”

Section: Literature Reviewmentioning

confidence: 99%

“…The workflow starts with signal filtering and QRS and R-peak detection algorithms methods, followed by feature extraction and simple classification with a classifier or classification fusion methods with multiple classifiers. A broad set of handcrafted features for ECG analysis, such as temporal relationships between waves, morphological descriptors, state-space features, linear transform, spectral representation, wavelet analysis, etc., have been described as well [11,12,[18][19][20][21][22][23][24]. Among the temporal features, a wide assortment of QRS morphological descriptors was mentioned, including QRS width, positive and negative peak amplitudes, QRS slopes, and cardiogram vector descriptors.…”

Section: Literature Reviewmentioning

confidence: 99%

See 4 more Smart Citations

Human-Assisted vs. Deep Learning Feature Extraction: An Evaluation of ECG Features Extraction Methods for Arrhythmia Classification Using Machine Learning

et al. 2022

View full text Add to dashboard Cite

The success of arrhythmia classification tasks with Machine Learning (ML) algorithms is based on the handcrafting extraction of features from Electrocardiography (ECG) signals. However, feature extraction is a time-consuming trial-and-error approach. Deep Neural Network (DNN) algorithms bypass the process of handcrafting feature extraction since the algorithm extracts the features automatically in their hidden layers. However, it is important to have access to a balanced dataset for algorithm training. In this exploratory research study, we will compare the evaluation metrics among Convolutional Neural Networks (1D-CNN) and Support Vector Machines (SVM) using a dataset based on the merged public ECG signals database TNMG and CINC17 databases. Results: Both algorithms showed good performance using the new, merged ECG database. For evaluation metrics, the 1D-CNN algorithm has a precision of 93.04%, an accuracy of 93.07%, a recall of 93.20%, and an F1-score of 93.05%. The SVM classifier (λ = 10, C = 10e9) achieved the best classification metrics with two combined, handcrafted feature extraction methods: Wavelet transforms and R-peak Interval features, which achieved an overall precision of 89.04%, accuracy of 92.00%, recall of 94.20%, and F1-score of 91.54%. As an unique input feature and SVM (λ=10,C=100), wavelet transforms achieved precision, accuracy, recall, and F1-score metrics of 86.15%, 85.33%, 81.16%, and 83.58%. Conclusion: Researchers face a challenge in finding a broad dataset to evaluate ML models. One way to solve this problem, especially for deep learning models, is to combine several public datasets to increase the amount of data. The SVM and 1D-CNN algorithms showed positive results with the merge of databases, showing similar F1-score, precision, and recall during arrhythmia classification. Despite the favorable results for both of them, it should be considered that in the SVM, feature selection is a time-consuming trial-and-error process; meanwhile, CNN algorithms can reduce the workload significantly. The disadvantage of CNN algorithms is that it has a higher computational processing cost; moreover, in the absence of access to powerful computational processing, the SVM can be a reliable solution.

show abstract

Section: Svm Classifier Vs 1d-cnn Algorithmmentioning

confidence: 90%