Six-center assessment of CNN-transformer with belief matching loss for patient-independent seizure detection in EEG
Neurologists typically identify epileptic seizures from electroencephalograms (EEGs) by visual inspection. This process is often time-consuming, especially for EEG recordings that last hours or days. To expedite the process, a reliable, automated, and patient-independent seizure detector is essentia...
| Main Authors: | , , , , , |
|---|---|
| Other Authors: | |
| Format: | Journal Article |
| Language: | English |
| Published: |
2023
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/169959 |
| _version_ | 1826128853150466048 |
|---|---|
| author | Peh, Wei Yan Thangavel, Prasanth Yao, Yuanyuan Thomas, John Tan, Yee-Leng Dauwels, Justin |
| author2 | Interdisciplinary Graduate School (IGS) |
| author_facet | Interdisciplinary Graduate School (IGS) Peh, Wei Yan Thangavel, Prasanth Yao, Yuanyuan Thomas, John Tan, Yee-Leng Dauwels, Justin |
| author_sort | Peh, Wei Yan |
| collection | NTU |
| description | Neurologists typically identify epileptic seizures from electroencephalograms (EEGs) by visual inspection. This process is often time-consuming, especially for EEG recordings that last hours or days. To expedite the process, a reliable, automated, and patient-independent seizure detector is essential. However, developing a patient-independent seizure detector is challenging as seizures exhibit diverse characteristics across patients and recording devices. In this study, we propose a patient-independent seizure detector to automatically detect seizures in both scalp EEG and intracranial EEG (iEEG). First, we deploy a convolutional neural network with transformers and belief matching loss to detect seizures in single-channel EEG segments. Next, we extract regional features from the channel-level outputs to detect seizures in multi-channel EEG segments. At last, we apply post-processing filters to the segment-level outputs to determine seizures' start and end points in multi-channel EEGs. Finally, we introduce the minimum overlap evaluation scoring as an evaluation metric that accounts for minimum overlap between the detection and seizure, improving upon existing assessment metrics. We trained the seizure detector on the Temple University Hospital Seizure (TUH-SZ) dataset and evaluated it on five independent EEG datasets. We evaluate the systems with the following metrics: sensitivity (SEN), precision (PRE), and average and median false positive rate per hour (aFPR/h and mFPR/h). Across four adult scalp EEG and iEEG datasets, we obtained SEN of 0.617-1.00, PRE of 0.534-1.00, aFPR/h of 0.425-2.002, and mFPR/h of 0-1.003. The proposed seizure detector can detect seizures in adult EEGs and takes less than 15[Formula: see text]s for a 30[Formula: see text]min EEG. Hence, this system could aid clinicians in reliably identifying seizures expeditiously, allocating more time for devising proper treatment. |
| first_indexed | 2024-10-01T07:31:19Z |
| format | Journal Article |
| id | ntu-10356/169959 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T07:31:19Z |
| publishDate | 2023 |
| record_format | dspace |
| spelling | ntu-10356/1699592023-08-16T04:48:49Z Six-center assessment of CNN-transformer with belief matching loss for patient-independent seizure detection in EEG Peh, Wei Yan Thangavel, Prasanth Yao, Yuanyuan Thomas, John Tan, Yee-Leng Dauwels, Justin Interdisciplinary Graduate School (IGS) Science::Medicine Transformer Belief Matching Neurologists typically identify epileptic seizures from electroencephalograms (EEGs) by visual inspection. This process is often time-consuming, especially for EEG recordings that last hours or days. To expedite the process, a reliable, automated, and patient-independent seizure detector is essential. However, developing a patient-independent seizure detector is challenging as seizures exhibit diverse characteristics across patients and recording devices. In this study, we propose a patient-independent seizure detector to automatically detect seizures in both scalp EEG and intracranial EEG (iEEG). First, we deploy a convolutional neural network with transformers and belief matching loss to detect seizures in single-channel EEG segments. Next, we extract regional features from the channel-level outputs to detect seizures in multi-channel EEG segments. At last, we apply post-processing filters to the segment-level outputs to determine seizures' start and end points in multi-channel EEGs. Finally, we introduce the minimum overlap evaluation scoring as an evaluation metric that accounts for minimum overlap between the detection and seizure, improving upon existing assessment metrics. We trained the seizure detector on the Temple University Hospital Seizure (TUH-SZ) dataset and evaluated it on five independent EEG datasets. We evaluate the systems with the following metrics: sensitivity (SEN), precision (PRE), and average and median false positive rate per hour (aFPR/h and mFPR/h). Across four adult scalp EEG and iEEG datasets, we obtained SEN of 0.617-1.00, PRE of 0.534-1.00, aFPR/h of 0.425-2.002, and mFPR/h of 0-1.003. The proposed seizure detector can detect seizures in adult EEGs and takes less than 15[Formula: see text]s for a 30[Formula: see text]min EEG. Hence, this system could aid clinicians in reliably identifying seizures expeditiously, allocating more time for devising proper treatment. 2023-08-16T04:48:48Z 2023-08-16T04:48:48Z 2023 Journal Article Peh, W. Y., Thangavel, P., Yao, Y., Thomas, J., Tan, Y. & Dauwels, J. (2023). Six-center assessment of CNN-transformer with belief matching loss for patient-independent seizure detection in EEG. International Journal of Neural Systems, 33(3), 2350012-. https://dx.doi.org/10.1142/S0129065723500120 0129-0657 https://hdl.handle.net/10356/169959 10.1142/S0129065723500120 36809996 2-s2.0-85149258776 3 33 2350012 en International Journal of Neural Systems © 2023 World Scientific Publishing Company. All rights reserved. |
| spellingShingle | Science::Medicine Transformer Belief Matching Peh, Wei Yan Thangavel, Prasanth Yao, Yuanyuan Thomas, John Tan, Yee-Leng Dauwels, Justin Six-center assessment of CNN-transformer with belief matching loss for patient-independent seizure detection in EEG |
| title | Six-center assessment of CNN-transformer with belief matching loss for patient-independent seizure detection in EEG |
| title_full | Six-center assessment of CNN-transformer with belief matching loss for patient-independent seizure detection in EEG |
| title_fullStr | Six-center assessment of CNN-transformer with belief matching loss for patient-independent seizure detection in EEG |
| title_full_unstemmed | Six-center assessment of CNN-transformer with belief matching loss for patient-independent seizure detection in EEG |
| title_short | Six-center assessment of CNN-transformer with belief matching loss for patient-independent seizure detection in EEG |
| title_sort | six center assessment of cnn transformer with belief matching loss for patient independent seizure detection in eeg |
| topic | Science::Medicine Transformer Belief Matching |
| url | https://hdl.handle.net/10356/169959 |
| work_keys_str_mv | AT pehweiyan sixcenterassessmentofcnntransformerwithbeliefmatchinglossforpatientindependentseizuredetectionineeg AT thangavelprasanth sixcenterassessmentofcnntransformerwithbeliefmatchinglossforpatientindependentseizuredetectionineeg AT yaoyuanyuan sixcenterassessmentofcnntransformerwithbeliefmatchinglossforpatientindependentseizuredetectionineeg AT thomasjohn sixcenterassessmentofcnntransformerwithbeliefmatchinglossforpatientindependentseizuredetectionineeg AT tanyeeleng sixcenterassessmentofcnntransformerwithbeliefmatchinglossforpatientindependentseizuredetectionineeg AT dauwelsjustin sixcenterassessmentofcnntransformerwithbeliefmatchinglossforpatientindependentseizuredetectionineeg |