Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2510.15717 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918192978329600 |
|---|---|
| author | Gashti, Mehdi Zekriyapanah Mohammadpour, Mostafa Eshkiki, Hassan Ghanbarizadeh, Vahid |
| author_facet | Gashti, Mehdi Zekriyapanah Mohammadpour, Mostafa Eshkiki, Hassan Ghanbarizadeh, Vahid |
| contents | Epileptic biomarkers play a crucial role in identifying the origin of seizures, an essential aspect of pre-surgical planning for epilepsy treatment. These biomarkers can vary significantly over time. By studying these temporal fluctuations, we can enhance their effectiveness in guiding surgical planning. This research focuses on examining how circadian rhythms influence epilepsy biomarkers and aims to determine the optimal times for their analysis. To investigate the relationship between epilepsy biomarkers and circadian rhythm, the sleep/wake states first need to be classified. After the biomarkers are identified, they are compared across these states. A retrospective analysis was conducted on intracranial electroencephalography data from patients with focal epilepsy. The biomarkers-spikes, spike sequences, high-frequency oscillations (HFOs), and pathological HFOs-were identified through automatic detection. The alpha/delta ratio was also calculated to distinguish between asleep and awake stages. Data from 9 patients were analyzed, and the classification of sleep and wake states was achieved with an area under the curve of 84%. All biomarker rates were higher during the sleep stage compared to the wake stage. Pathological HFOs and the sequence of spikes proved to be more precise indicators regarding distance to seizure onset than spikes or HFOs. Unlike previous studies that relied predominantly on long-term spike biomarker analysis, this study is the first to utilize a comprehensive set of biomarkers, including HFOs, spike sequences, and pathological HFOs, to enhance seizure onset zone prediction. The rates of epilepsy biomarkers during sleep vary considerably from those seen while awake, making sleep data analysis more effective in accurately predicting the seizure onset zone. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_15717 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Automated Detection of Circadian-Dependent Epileptic Biomarkers for Seizure Localization: A Machine Learning and Signal Processing Framework Gashti, Mehdi Zekriyapanah Mohammadpour, Mostafa Eshkiki, Hassan Ghanbarizadeh, Vahid Signal Processing 68T05, 92C55 I.5.1; I.2.6 Epileptic biomarkers play a crucial role in identifying the origin of seizures, an essential aspect of pre-surgical planning for epilepsy treatment. These biomarkers can vary significantly over time. By studying these temporal fluctuations, we can enhance their effectiveness in guiding surgical planning. This research focuses on examining how circadian rhythms influence epilepsy biomarkers and aims to determine the optimal times for their analysis. To investigate the relationship between epilepsy biomarkers and circadian rhythm, the sleep/wake states first need to be classified. After the biomarkers are identified, they are compared across these states. A retrospective analysis was conducted on intracranial electroencephalography data from patients with focal epilepsy. The biomarkers-spikes, spike sequences, high-frequency oscillations (HFOs), and pathological HFOs-were identified through automatic detection. The alpha/delta ratio was also calculated to distinguish between asleep and awake stages. Data from 9 patients were analyzed, and the classification of sleep and wake states was achieved with an area under the curve of 84%. All biomarker rates were higher during the sleep stage compared to the wake stage. Pathological HFOs and the sequence of spikes proved to be more precise indicators regarding distance to seizure onset than spikes or HFOs. Unlike previous studies that relied predominantly on long-term spike biomarker analysis, this study is the first to utilize a comprehensive set of biomarkers, including HFOs, spike sequences, and pathological HFOs, to enhance seizure onset zone prediction. The rates of epilepsy biomarkers during sleep vary considerably from those seen while awake, making sleep data analysis more effective in accurately predicting the seizure onset zone. |
| title | Automated Detection of Circadian-Dependent Epileptic Biomarkers for Seizure Localization: A Machine Learning and Signal Processing Framework |
| topic | Signal Processing 68T05, 92C55 I.5.1; I.2.6 |
| url | https://arxiv.org/abs/2510.15717 |