Automated Deep Neural Network Approach for Detection of Epileptic Seizures
Bibliographic record
Abstract
In this thesis, I focus on exploiting electroencephalography (EEG) signals for early seizure diagnosis in patients. This process is based on a powerful deep learning algorithm for times series data called Long Short-Term Memory (LSTM) network. Since manual and visual inspection (detection) of epileptic seizure through the electroencephalography (EEG) signal by expert neurologists is time-consuming, work-intensive and error-prone and it might take a couple hours for experts to analyze a single patient record and to do recognition when immediate action is needed to be taken. This thesis proposes a reliable automatic seizure/non-seizure classification method that could facilitate the identification process of characteristic epileptic patterns, such as pre-ictal spikes, seizures and determination of seizure frequency, seizure type, etc. In order to recognize epileptic seizure accurately, the proposed model exploits the temporal dependencies in the EEG data. Experiments on clinical data present that this method achieves a high seizure prediction accuracy and maintains reliable performance. This thesis also finds the most efficient lengths of EEG recording for highest accuracies of different classification in the automated seizure detection realm. It could help non-experts to predict the seizure more comprehensively and bring awareness to patients and caregivers of upcoming seizures, enhancing the daily lives of patients against unpredictable occurrence of seizures.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".