Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

Fig 1.

ECG signal compression classification process.

More »

Fig 1 Expand

Fig 2.

(a-d) ECG plotted for the first 1000 data of the subject.

More »

Fig 2 Expand

Fig 3.

Schematic diagram of heartbeat beat interception centered on the R-peak.

More »

Fig 3 Expand

Table 1.

Number of each heartbeat beat category after cutting and the actual number used [49].

More »

Table 1 Expand

Fig 4.

Proposed compression process.

More »

Fig 4 Expand

Table 2.

Detailed parameters of the compression network at CR = 0.2.

More »

Table 2 Expand

Fig 5.

Proposed classification network.

More »

Fig 5 Expand

Table 3.

Detailed parameters of the classification network at CR = 0.2.

More »

Table 3 Expand

Fig 6.

Variation of model learning rate with epoch at CR = 0.1.

More »

Fig 6 Expand

Fig 7.

Training and validation performances using a proposed model with ECG datasets.

More »

Fig 7 Expand

Table 4.

Accuracy of different compression methods in the MIT-BIH arrhythmia database test set at different compression ratios.

More »

Table 4 Expand

Fig 8.

Accuracy trends in the MIT-BIH arrhythmia database using different compression methods at different compression ratios.

More »

Fig 8 Expand

Fig 9.

Confusion matrix of the proposed model on the MIT-BIH arrhythmia database test set at CR = 0.5.

More »

Fig 9 Expand

Fig 10.

Trends in accuracy using different classification methods at different compression ratios in the MIT-BIH arrhythmia database.

More »

Fig 10 Expand

Fig 11.

ROC curve of each classification model when CR = 0.05.

More »

Fig 11 Expand

Fig 12.

PR curve of each classification model when CR = 0.05.

More »

Fig 12 Expand

Table 5.

Accuracy of different classification methods in the MIT-BIH arrhythmia database test set at different compression ratios.

More »

Table 5 Expand

Table 6.

The time (s) spent in parameter training for each classification network.

More »

Table 6 Expand

Table 7.

Average precision, sensitivity and F1-score of each classification network at CR = 0.4.

More »

Table 7 Expand

Table 8.

Average precision, sensitivity and F1-score of each classification network at CR = 0.2.

More »

Table 8 Expand

Fig 13.

Accuracy trends using different compression methods at different compression ratios in the TianChi ECG signal database.

More »

Fig 13 Expand

Fig 14.

Confusion matrix of the proposed model on the TianChi ECG signal database test set at CR = 0.1.

More »

Fig 14 Expand

Table 9.

Accuracy of different compression methods in TianChi ECG signal database test set at different compression ratios.

More »

Table 9 Expand

Fig 15.

Accuracy trends using different classification methods at different compression ratios in the TianChi ECG signal database.

More »

Fig 15 Expand

Table 10.

Accuracy of different classification methods in the TianChi ECG signal database test set at different compression ratios.

More »

Table 10 Expand

Fig 16.

Accuracy trends using different classification methods at different compression ratios in the MIT-BIH ECG signal database.

More »

Fig 16 Expand

Table 11.

Accuracy of different classification methods in the MIT-BIH ECG signal database test set at different compression ratios.

More »

Table 11 Expand