Skip to main content
. Author manuscript; available in PMC: 2021 Feb 10.
Published in final edited form as: Comput Biol Med. 2020 Oct 15;127:104057. doi: 10.1016/j.compbiomed.2020.104057

Table 2.

Performance of the proposed model for the AF classification task on the MIT-BIH AFIB database while the database is not balanced and the data segments for the training and test phases come from different ECG recordings.

Method Database Best Performance (%)
Sensitivity Specificity Accuracy AUC
HAN-ECG3 AFDB 90.53 79.54 82.41 89.46
HAN-ECG2 AFDB 89.86 77.49 81.58 88.65
HAN-ECG1 AFDB 89.47 75.15 79.96 85.94
RNN AFDB 89.20 74.38 79.55 85.88