Abstract:Our study concerns with automated predicting of congestive heart failure (CHF) through the analysis of electrocardiography (ECG) signals. A novel machine learning approach, regularized hessenberg decomposition based extreme learning machine (R-HessELM), and feature models; squared, circled, inclined and grid entropy measurement were introduced and used for prediction of CHF. This study proved that inclined entropy measurements features well represent characteristics of ECG signals and together with R-HessELM approach overall accuracy of 98.49% was achieved.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to automatically predict congestive heart failure (CHF) by analyzing electrocardiogram (ECG) signals. Specifically, the authors introduced a new machine - learning method - Regularized Hessenberg Decomposition - based Extreme Learning Machine (R - HessELM), as well as several new feature models, such as Squared Entropy Measurement (SEM), Circular Entropy Measurement (CEM), Inclined Entropy Measurement (IEM) and Grid Entropy Measurement (GEM), to improve the accuracy of CHF prediction. ### Key issues 1. **Automated CHF prediction**: CHF is a serious medical condition, and early diagnosis is crucial for improving patients' treatment outcomes and quality of life. Traditional clinical techniques have difficulty in easily assessing the presence of CHF, so it is necessary to develop a reliable automated prediction system. 2. **Improving prediction accuracy**: Existing methods have certain limitations in predicting CHF, especially when dealing with complex ECG signals. The new method proposed in this paper aims to improve prediction accuracy by introducing new feature extraction techniques and an improved machine - learning algorithm. ### Method overview - **Data sources**: Use the CHF database and the normal sinus rhythm database on the Physionet website. - **Feature extraction**: Extract multiple features from ECG signals, especially calculate entropy measurement values in different regions through SODP (Second - Order Difference Plot). - **Classifier**: Use the Regularized Hessenberg Decomposition - based Extreme Learning Machine (R - HessELM) for classification. - **Performance evaluation**: Evaluate the performance of the model through methods such as cross - validation and Leave - One - Out (LOO). ### Main contributions - **New feature models**: Introduced four new entropy measurement methods (SEM, CEM, IEM, GEM), which can better capture the features in ECG signals. - **Improved machine - learning algorithm**: Proposed an Extreme Learning Machine based on Hessenberg Decomposition and further optimized the generalization ability of the model through regularization. - **High accuracy**: Experimental results show that the overall accuracy rate using the IEM feature and the R - HessELM model reached 98.49%. ### Conclusion This study successfully developed an efficient automated CHF prediction system, which can reduce the possibility of misdiagnosis while maintaining high accuracy. This provides strong support for future medical applications, especially in early diagnosis and prevention.

Regularized HessELM and Inclined Entropy Measurement for Congestive Heart Failure Prediction

Entropy-based reliable non-invasive detection of coronary microvascular dysfunction using machine learning algorithm

High-Performance Personalized Heartbeat Classification Model for Long-Term ECG Signal

Prediction of Left Ventricular Ejection Fraction Changes in Heart Failure Patients Using Machine Learning and Electronic Health Records: a Multi-Site Study.

Application of empirical mode decomposition (EMD) for automated identification of congestive heart failure using heart rate signals

Prediction of heart failure patients with distinct left ventricular ejection fraction levels using circadian ECG features and machine learning

Large Language Model-informed ECG Dual Attention Network for Heart Failure Risk Prediction

Application of Deep Neural Network for Congestive Heart Failure Detection Using ECG Signals

Interpretable estimation of the risk of heart failure hospitalization from a 30-second electrocardiogram

[Atypical depression at the onset of endogenous diseases in young adults (clinical picture and differential diagnosis)].

Model for classification of heart failure severity in patients with hypertrophic cardiomyopathy using a deep neural network algorithm with a 12-lead electrocardiogram

Time-domain heart rate variability features for automatic congestive heart failure prediction

Efficient Extreme Learning Machine (ELM) Based Algorithm for Electrocardiogram (ECG) Heartbeat Classification

An Integrated Machine Learning Approach for Congestive Heart Failure Prediction

Non-Invasive Heart Failure Evaluation Using Machine Learning Algorithms

Patient Specific Congestive Heart Failure Detection From Raw ECG signal

Empirical Mode Decomposition as a Novel Approach to Study Heart Rate Variability in Congestive Heart Failure Assessment

Inter-Patient Congestive Heart Failure Detection Using ECG-Convolution-Vision Transformer Network

Understanding Heart-Failure Patients EHR Clinical Features via SHAP Interpretation of Tree-Based Machine Learning Model Predictions

Score and Correlation Coefficient-Based Feature Selection for Predicting Heart Failure Diagnosis by Using Machine Learning Algorithms

A New Approach for Congestive Heart Failure and Arrhythmia Classification Using Angle Transformation with LSTM