Table 1 CNN based model classification results, evaluated for each participant(testing dataset and training dataset).
Subject id | Sensitivity % | Specificity % | Accuracy % | Number of correctly predicted 10 minutes intervals/total | |||
|---|---|---|---|---|---|---|---|
Individual beat | 10 min voting | Individual beat | 10 min voting | Individual beat | 10 min voting | ||
NN based system classification results, evaluated on testing data for each participant | |||||||
Subject 1 | 74.2 | 78.0 | 71.2 | 77.1 | 72.3 | 77.4 | 106/141 = 75.2% |
Subject 2 | 66.0 | 79.8 | 69.5 | 77.1 | 69.3 | 77.3 | 146/187 = 78.1% |
Subject 3 | 82.2 | 100 | 87.4 | 91.9 | 87.1 | 92.4 | 168/183 = 91.8% |
Subject 4 | 81.1 | 91.5 | 76.3 | 80.5 | 77.2 | 82.6 | 128/156 = 82.1% |
Average | 75.9 ± 7.4 | 87.5 ± 10.3 | 76.1 ± 8.0 | 81.7 ± 7.0 | 76.5 ± 7.7 | 82.4 ± 7.0 | 81.8% |
CNN based system classification results, evaluated on training data for each participant. | |||||||
Subject 1 | 91.3 | 94.6 | 79.7 | 80.4 | 84.8 | 86.5 | 118/136 = 86.8% |
Subject 2 | 93.1 | 100 | 87.5 | 93.2 | 88.5 | 94.5 | 174/184 = 94.6% |
Subject 3 | 97.5 | 100 | 88.4 | 89.7 | 89.9 | 91.5 | 247/270 = 91.5% |
Subject 4 | 75.0 | 83.5 | 81.3 | 85.5 | 78.3 | 84.6 | 138/163 = 84.7% |
Average | 89.2 ± 9.8 | 94.5 ± 7.7 | 84.2 ± 4.3 | 87.2 ± 5.5 | 85.4 ± 5.1 | 89.3 ± 4.5 | 89.4% |
