Text
IMPLEMENTASI DISCRETE FOURIER TRANSFORM DAN ARSITEKTUR WAVENET – LONG SHORT-TERM MEMORY PADA KLASIFIKASI BEAT SINYAL ELEKTROKARDIOGRAM
In electrocardiogram (ECG) signal recordings, heartbeat rhythms can be normal and abnormal. One way to identify heart abnormalities in ECG signals is by classifying heartbeats into five classes, namely non ectopic beat (N), supraventricular ectopic beat (S), ventricular ectopic beat (V), fusion beat (F), and unknown beat (Q). This study uses a combination of WaveNet architecture and Long Short-Term Memory (LSTM) for ECG signal classification, with the preprocessing stage using the Discrete Fourier Transform (DFT) method. The DFT method obtained an average SNR value of 25.21 dB which falls into the good category. This shows that the signal produced is good with low noise. Evaluation of model performance is done by measuring accuracy, sensitivity, specificity, precision, and F1-score. The accuracy value obtained is 99.60%, indicating that the model is able to classify the ECG signal into each class very well. Precision obtained 99.134% indicates that the model has a good level of accuracy in predicting arrhythmia disease for each class, so there are only a few classification errors. Sensitivity was 99.124%, indicating the model's good ability to recognize data belonging to the five classes. Specificity obtained 99.78% indicates that the model can recognize data that is not part of the five classes. F1-score obtained 99.134% indicates a balance between precision and sensitivity. Based on the evaluation results, the model shows excellent performance in recognizing class F compared to classes N, Q, S, and V with a sensitivity value of 100%. The values obtained in classes N, Q, S, and V are still below class F, but the results obtained are very good because the sensitivity values for classes S, V, and Q are more than 90%. Overall, the proposed model has shown excellent performance.
Inventory Code | Barcode | Call Number | Location | Status |
---|---|---|---|---|
2507001896 | T169738 | T1697382025 | Central Library (Reference) | Available but not for loan - Not for Loan |
No other version available