Fig 1.
13 MFFCs are extracted from each interval (i.e., S1, systole, S2, and diastole) of a given PCG beat, summing up to 52 MFCCs for that beat.
Fig 2.
A: Single-classifier strategy: The MFCCs extracted from the first 9 beats of a PCG are averaged and fed to a single classifier, B: Ensemble-classifier strategy: Nine classifiers are used separately to distinguish normal beats from abnormal ones. In the end, if the number of the normal beats is more than the abnormal beats, the PCG signal is decided to be normal; otherwise, it is abnormal.
Fig 3.
The staircase graph shows the segmentation results of the plotted PCG signal: Level 1 shows S1 intervals, level 2 shows systole intervals, level 3 shows S2 intervals, and level 4 shows diastole intervals.
Table 1.
The results for the single-classifier strategy for the three types of classifiers: kNN (k = 3), SVM with a polynomial kernel, and DT.
Table 2.
The results for the ensemble classification strategy for three types of classifiers: kNN (k = 3), SVM with a polynomial kernel, and DT.
Fig 4.
The classification accuracy for the single- (light gray) and the ensemble classifiers (dark gray).
The ensemble classifier has outperformed the single classifier for both the decision tree (DT) and the SVM classifier types. The error bars represent the 95% confidence interval for the classification accuracy.
Fig 5.
The classification sensitivity for the single- (light gray) and the ensemble classifiers (dark gray).
The ensemble classifier has outperformed the single classifier for both the decision tree (DT) and the SVM classifier types. The error bars represent the 95% confidence interval for the classification sensitivity.
Fig 6.
The classification specificity for the single- (light gray) and the ensemble classifiers (dark gray).
The ensemble classifier has outperformed the single classifier only for the decision tree (DT) classifier type. The error bars represent the 95% confidence interval for the classification specificity.
Table 3.
Comparison of the results of this study with similar studies.
Fig 7.
Comparison of the accuracy, sensitivity, and specificity of our ensemble and single classifiers with the some of the methods cited in Table 3.