Key result
Improved UNet++ model achieves ~89% accuracy in detecting CHF using short-term RR intervals.
Why the study?
Congestive heart failure is difficult to detect at an early stage, prompting interest in heart rate variability as a prognostic indicator.
Does an improved UNet++ model using short-term RR intervals improve diagnostic accuracy for congestive heart failure compared to state-of-the-art methods?
Does an improved UNet++ model using short-term RR intervals improve diagnostic accuracy for congestive heart failure compared to state-of-the-art methods?
Effect estimate: Accuracy 88.79%
A deep learning model based on UNet++ using short-term heart rate variability signals demonstrated high accuracy for the detection of congestive heart failure.
May support DL-based CHF screening from short-term RR intervals; leaves open clinical utility pending prospective validation.
Congestive heart failure (CHF), a progressive and complex syndrome caused by ventricular dysfunction, is difficult to detect at an early stage. Heart rate variability (HRV) was proposed as a prognostic indicator for CHF. Inspired by the success of 2-D UNet++ in medical image segmentation, in this paper, we introduce an end-to-end encoder-decoder model to detect CHF using HRV signals. The developed model enhances the UNet++ model with Squeeze-and-Excitation (SE) residual blocks to extract deep features hierarchically and distinguish CHF patients from normal subjects. Two open-source databases are utilized for evaluating the proposed method, and three segment lengths of intervals between successive R-peaks are employed in comparison with state-of-the-art methods. The proposed method achieves an accuracy of 85.64%, 86.65% and 88.79% when 500, 1000 and 2000 RR intervals are utilized, respectively. It demonstrates that HRV evaluation based on deep learning can be an important tool for early detection of CHF, and may assist clinicians in achieving timely and accurate diagnoses.
No takes yet. Share an insight, caveat, or question.
Lei et al. (2021) studied Congestive heart failure (n=83). Improved 1-D UNet++ model with Squeeze-and-Excitation residual blocks vs. Other machine learning models / Normal subjects was evaluated on Diagnostic accuracy using 2000 short-term RR intervals (Accuracy 88.79%). The improved 1-D UNet++ model with Squeeze-and-Excitation residual blocks achieved a diagnostic accuracy of 88.79% in detecting congestive heart failure using 2000 short-term RR intervals.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: