Improving Accuracy of Heart Failure Detection Using Data Refinement

Abstract

Due to the wide inter- and intra-individual variability, short-term heart rate variability (HRV) analysis (usually 5 min) might lead to inaccuracy in detecting heart failure. Therefore, RR interval segmentation, which can reflect the individual heart condition, has been a key research challenge for accurate detection of heart failure. Previous studies mainly focus on analyzing the entire 24-h ECG recordings from all individuals in the database which often led to poor detection rate. In this study, we propose a set of data refinement procedures, which can automatically extract heart failure segments and yield better detection of heart failure. The procedures roughly contain three steps: (1) select fast heart rate sequences, (2) apply dynamic time warping (DTW) measure to filter out dissimilar segments, and (3) pick out individuals with large numbers of segments preserved. A physical threshold-based Sample Entropy (SampEn) was applied to distinguish congestive heart failure (CHF) subjects from normal sinus rhythm (NSR) ones, and results using the traditional threshold were also discussed. Experiment on the PhysioNet/MIT RR Interval Databases showed that in SampEn analysis (embedding dimension m = 1, tolerance threshold r = 12 ms and time series length N = 300), the accuracy value after data refinement has increased to 90.46% from 75.07%. Meanwhile, for the proposed procedures, the area under receiver operating characteristic curve (AUC) value has reached 95.73%, which outperforms the original method (i.e., without applying the proposed data refinement procedures) with AUC of 76.83%. The results have shown that our proposed data refinement procedures can significantly improve the accuracy in heart failure detection.

Correlation between Levels of Serum Lipoprotein-Associated Phospholipase A2 and Soluble Suppression of Tumorigenicity 2 and Condition of Acute Heart Failure Patients and Their Predictive Value for Prognosis

“…A large number of epidemiological trials have shown that LP-PLA2 is an independent factor for predicting CVE [ 22 ]. Scholars Xiong et al corrected the risk factors of traditional CVD and found that LP-PLA2 was still a risk predictor for CVE and had a positive correlation with the severity of CHF; e.g.,, the elevated level of CHF could increase the LP-PLA2 level [ 23 ]. In this study, no statistical differences in the enrolled patients' general information such as age and gender were observed ( P > 0.05), the serum LP-PLA2 levels were obviously different between the severe group and the mild group, and according to the logistic regression analysis, serum LP-PLA2 had independent predictive value for AHF prognosis, demonstrating that its level was of great importance in AHF and not just in CHD and CHF.…”

Entropy-Based Strategies for Rapid Pre-Processing and Classification of Time Series Data from Single-Molecule Force Experiments

“…While this method offers a well-defined information-based predictability assessment, it faces the fundamental problems that have arisen concerning data processing. Efforts for less demanding processing, especially of biologically relevant data, were subsequently reflected in the design of approximate entropy (ApEn) and later in the modified sample entropy variant (SampEn) [ 7 , 8 ]. Some of the shortcomings of these approaches have been addressed by the adjustments contained in the multiscale description (MSE) [ 9 ].…”