Skip to main content

Table 3 Comparison of classification results of CNN_Transformer_LC and baselines on different Arrhythmia classifications

From: Constrained transformer network for ECG signal processing and arrhythmia classification

Methods Normal AF I-AVF LBBB RBBB PAC PVC STD STE F1
CNN 0.578 0.709 0.753 0.773 0.825 0.207 0.376 0.562 0.389 0.574
ResNet 0.578 0.787 0.833 0.757 0.847 0.324 0.407 0.610 0.260 0.601
Mutil_channelCNN 0.666 0.733 0.827 0.8 0.821 0.421 0.648 0.575 0.32 0.646
BiRNN 0.738 0.768 0.742 0.705 0.821 0.59 0.807 0.658 0.294 0.742
CNN_BiLSTM 0.723 0.826 0.851 0.829 0.893 0.600 0.818 0.692 0.529 0.751
CNN_Transformer_LC 0.817 0.858 0.878 0.800 0.872 0.618 0.830 0.711 0.686 0.786
  1. Bold values indiate the best experimental results under this category