Robust Detection of Atrial Fibrillation Using Classification of a Linearly-Transformed Window of R-R Intervals Tachogram

Abstract

Atrial fibrillation (AF) is the most common cardiac arrhythmia. It increases the risk of stroke, dementia, and death; therefore, its timely diagnosis at an initial stage is crucial. Often wearable mobile devices are recommended for the primary detection of this life-threatening arrhythmia. Irregularity of the heartbeat duration, often measured through R-R intervals (RRI), has been intensively investigated during the past four decades for automatic detection of AF. However, little improvement has been made when the input signal (RRI tachogram) contains different types of arrhythmic rhythms. In this paper, we propose a neighborhood component analysis (NCA) based linear transformation of a window of RRI tachogram to improve the robustness of AF detection. Several state-of-the-art classification models are trained and tested using transformed signals, and AF detection performance are evaluated using the challenging MIT-BIH Arrhythmia Database containing various types of arrhythmic rhythms. The experimental results show significant improvement in AF detection performance using the transformed signals compared to those for signals in the original space and after linear-discriminant-analysis-based transformation. In particular, for the Naïve Bayesian classification of the transformed signals, we obtained 98.59% sensitivity, 99.91% specificity, 99.16% positive predictive value, and 99.79% accuracy. The proposed AF detection method outperforms the existing methods reported in the past four decades. Owing to the use of a short window of RRI tachogram (15 consecutive RRIs), the proposed method can be incorporated into a deployable mobile screening device for robust detection of AF.

Non-Contact Atrial Fibrillation Detection From Face Videos by Learning Systolic Peaks

“…The AF detection from ECG can be based on 1) P-wave detection [2]- [5], 2) R-R interval variability [5]- [9], and 3) deep learning (DL)-based methods [10]- [12]. P waves are not prominent features and are often influenced by artifacts, while R peaks have higher amplitudes that can be easily detected.…”

“…Some traditional methods [6]- [8] extracted features such as entropy [8] or HRV features [7] from R-R intervals and simply use a threshold to do AF classification. Some methods [9], [32], [33] applied machine learning methods such as support vector machine (SVM) [32], convolutional neural network [33] or neighborhood component analysis [9] to R-R intervals for AF detection and achieved higher performance than traditional methods. However, it is difficult for rhythm-based methods using R-R intervals to distinguish AF from other arrhythmias, while the morphology-based methods such as P-wave absence can avoid false-positive errors in AF detection [34].…”

“…Theoretically, the ECG and contact PPG results should be the upper bound that the model can achieve and are the reference results compared with remote PPG results. We also report results of ECG-based AF detection methods [9], [11] for healthy vs. AF and SR vs. AF. For remote PPG, we compare our results with four previous methods [28], [29], [38], [45].…”

“…For the baseline methods [28], [38], [45] and ECG/contact PPG reference methods, we use the same SVM classifier and HRV features from Section III-C, while only change the input IBI series. For [9], [11], [29], we use the features and models in their papers.…”

Multiscale Encoding of Electrocardiogram Signals with a Residual Network for the Detection of Atrial Fibrillation

“…With traditional machine learning, the process of AF detection starts with the extraction of handcrafted features and continues with the use of these features to train a classification model. Several classifiers were utilized, including support vector machines (SVMs) [ 8 , 9 , 10 ], k-nearest neighbors [ 11 , 12 ], and an artificial neural network (ANN) classifier [ 13 , 14 ]. The detection or classification performance of these techniques was significantly affected by the quality of the extracted features.…”