HKF: Hierarchical Kalman Filtering With Online Learned Evolution Priors for Adaptive ECG Denoising (2022-10-23T00:00:00.000000Z)

TL;DR

HKF, a hierarchical and adaptive Kalman filter, which uses a proprietary state space model to effectively capture both intra- and inter-heartbeat dynamics for ECG signal denoising, makes it a suitable candidate for applications in extramural healthcare settings.

Abstract

Electrocardiography (ECG) signals play a pivotal role in many healthcare applications, especially in at-home monitoring of vital signs. Wearable technologies, which these applications often depend upon, frequently produce low-quality ECG signals. While several methods exist for ECG denoising to enhance signal quality and aid clinical interpretation, they often underperform with ECG data from wearable technology due to limited noise tolerance or inadequate flexibility in capturing ECG dynamics. This paper introduces HKF, a hierarchical and adaptive Kalman filter, which uses a proprietary state space model to effectively capture both intra- and inter-heartbeat dynamics for ECG signal denoising. HKF learns a patient-specific structured prior for the ECG signal's intra-heartbeat dynamics in an online manner, resulting in a filter that adapts to the specific ECG signal characteristics of each patient. In an empirical study, HKF demonstrated superior denoising performance (reduced Mean-Squared Error) while preserving the unique properties of the waveform. In a comparative analysis, HKF outperformed previously proposed methods for ECG denoising, such as the model-based Kalman filter and data-driven autoencoders. This makes it a suitable candidate for applications in extramural healthcare settings.

Authors

Nir Shlezinger

2 papers

Timur Locher

1 papers

Guy Revach

1 papers

TL;DR

Abstract

Authors

References90 items

Adaptive Kalmannet: Data-Driven Kalman Filter with Fast Adaptation

Analysis of ECG-based arrhythmia detection system using machine learning

Review of ECG data feature processing and classification

Onset Detection of Pulse-Shaped Bioelectrical Signals Using Linear State Space Models

A deep learning-based framework For ECG signal denoising based on stacked cardiac cycle tensor

A Review on Computational Methods for Denoising and Detecting ECG Signals to Detect Cardiovascular Diseases

RTSNet: Learning to Smooth in Partially Known State-Space Models

ECG denoising and feature extraction techniques – a review

KalmanNet: Neural Network Aided Kalman Filtering for Partially Known Dynamics

Comprehensive comparison of Apple Watch and Fitbit monitors in a free-living setting

Signal Analysis Using Local Polynomial Approximations

An Efficient ECG Denoising Method Based on Empirical Mode Decomposition, Sample Entropy, and Improved Threshold Function

Multi-Channel Fetal ECG Denoising With Deep Convolutional Neural Networks

Review of noise removal techniques in ECG signals

Wearable Devices for Ambulatory Cardiac Monitoring: JACC State-of-the-Art Review.

Robust ECG R-peak detection using LSTM

An Adaptive Kalman Filter Bank for ECG Denoising

A New ECG Denoising Framework Using Generative Adversarial Network

An ECG Signal Prediction Method Based on ARIMA Model and DWT

Deep Learning Models for Denoising ECG Signals

Riemannian Geometry of Symmetric Positive Definite Matrices via Cholesky Decomposition

Noise Reduction in ECG Signals Using Fully Convolutional Denoising Autoencoders

Deep Recurrent Neural Networks for ECG Signal Denoising

Windowed State-Space Filters for Signal Detection and Separation

Mathematical model with autoregressive process for electrocardiogram signals

A Signal Decomposition Model-Based Bayesian Framework for ECG Components Separation

An Energy-Efficient Design for ECG Recording and R-Peak Detection Based on Wavelet Transform

Analysis and detection R-peak detection using Modified Pan-Tompkins algorithm

Conic Geometric Optimization on the Manifold of Positive Definite Matrices

Morphological modeling of cardiac signals based on signal decomposition

Bayesian Filtering and Smoothing

Empirical Wavelet Transform

ECG denoising using extended Kalman filter

Denoising of ECG signals based on noise reduction algorithms in EMD and wavelet domains

Stochastic Gradient Descent on Riemannian Manifolds

ST Analysis of the Fetal ECG, as an Adjunct to Fetal Heart Rate Monitoring in Labour: A Review.

An Adaptive Kalman Filter for ECG Signal Enhancement

Cyclostationary Signal Analysis

ECG denoising using modulus maxima of wavelet transform

Model-Based ECG Denoising Using Empirical Mode Decomposition

Expectation Maximization as Message Passing - Part I: Principles and Gaussian Messages

ECG Signal Denoising By Wavelet Transform Thresholding

ECG Denoising and Compression Using a Modified Extended Kalman Filter Structure

A novel approach in R peak detection using Hybrid Complex Wavelet (HCW).

A Nonlinear Bayesian Filtering Framework for ECG Denoising

Novel dry electrodes for ECG monitoring

Geometric Means in a Novel Vector Space Structure on Symmetric Positive-Definite Matrices

Real time electrocardiogram QRS detection using combined adaptive threshold

A dynamical model for generating synthetic electrocardiogram signals

Was it a heart attack?

Cardiotocography only versus cardiotocography plus ST analysis of fetal electrocardiogram for intrapartum fetal monitoring: a Swedish randomised controlled trial

The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis

Assessing ECG signal quality on a coronary care unit

Parameter estimation for linear dynamical systems

Origin of the electrocardiogram

The discrete wavelet transform: wedding the a trous and Mallat algorithms

Ten Lectures on Wavelets

Fetal acidaemia, the cardiotocograph and the T/QRS ratio of the fetal ECG in labour

The likelihood for a state space model

[Long Q-T syndrome].

Changes in the ST waveform of the fetal lamb electrocardiogram with hypoxemia.

QT Interval Prolongation as Predictor of Sudden Death in Patients with Myocardial Infarction

On the maximum likelihood estimates for linear dynamic systems

ARRHYTHMIA

“Hierarchical ﬁltering with online learned priors for ECG denoising,”

Localized state space and polynomial filters with applications in electrocardiography

Denoising of ECG signal by non-local estimation of approximation coefficients in DWT

Noise Analysis and Different Denoising Techniques of ECG Signal - A Survey

Deep Learning

Denoising of Electrocardiogram Data with Methods of Wavelet Transform

Real Time Electrocardiogram Segmentation for Finger based ECG Biometrics

Orfanidis, Applied Optimum Signal Processing

Positive Deﬁnite Matrices (Princeton Ser. Appl. Math.). Princeton, NJ, USA

ECG Denoising Based on the Empirical Mode Decomposition

ECG Denoising with Adaptive Bionic Wavelet Transform

Electrocardiogram