Table 11 Comparison of findings based on the Bonn and New Delhi dataset
From: A hybrid CNN-Bi-LSTM model with feature fusion for accurate epilepsy seizure detection
Reference | Year | Strategies | Case | Accuracy (%) |
|---|---|---|---|---|
[58] | 2018 | EMD + Hilbert Transform + SVM | A-D-E AB-CD-E | 85.00 83.00 |
[59] | 2021 | Cov–Det + KST-MWUT + AB–BP–NN | C-E D-E AB-E CD-E ACD-E ABCD-E | 98.50 99.00 98.00 98.20 98.00 98.50 |
[35] | 2021 | TQWT + (Statistical + Frequency + Fractal and Entropy Features) + CNN–RNN | C-E D-E | 99.51 99.82 |
[60] | 2022 | DWT + InfoGain and Variance + FRNN | C-E D-E CD-E | 99.67 99.50 98.00 |
[41] | 2023 | DWT + Entropy Features + RF + CNN | A-E B-E AD-E BD-E ABC-E ABD-E BCD-E Interictal-Ictal Preictal-Ictal Non-ictal-Ictal | 99.30 98.10 99.28 97.46 98.95 97.30 97.65 100.00 97.33 98.33 |
[45] | 2023 | CNN-LSTM | C-E D-E AB-E CD-E ABCD-E | 98.20 97.60 98.30 97.90 98.70 |
[61] | 2024 | TCN-SA | A-E B-E | 97.37 93.50 |
Proposed Method | 2024 | DWT + (Time domain + Non-linear Features) + SVM-REF + CNN-Bi-LSTM | A-E B-E C-E D-E AB-E AC-E AD-E BC-E BD-E CD-E ABC-E ABD-E ACD-E BCD-E ABCD-E A-D-E AB-CD-E Interictal-Ictal Preictal-Ictal Non-ictal-Ictal | 99.50 98.17 99.75 100.00 98.60 99.33 99.61 98.28 97.94 99.11 99.08 98.17 99.50 98.08 98.93 96.19 95.71 100.00 98.83 99.67 |