Pre-Ictal Window Identification via Logistic Regression Open Access
Duggireddy, Hithardhi (Spring 2020)
Published
Abstract
Epilepsy is a debilitating disease characterized by spontaneous recurring seizures. It is the 4th most common neurological disorder. Mesial temporal lobe epilepsy (MTLE) is the most prevalent form of epilepsy and is the most resistant to medical therapy, with over 30% of patients failing to achieve adequate seizure control with antiepileptics. These patients may require surgery, but, resection of the epileptic focus is often not possible. An alternative therapy for these patients may be neural modulation via electrical stimulation; however, patients currently only experience approximately a 50% reduction in seizure frequency. A better fundamental understanding of mechanisms underlying seizure transition may be necessary for neural modulation to become more effective as a prophylactic therapeutic. Currently, research is being done to develop seizure prediction models to aid neural modulation; however, the pre-ictal windows used to develop these models are chosen arbitrarily. Therefore, the objective of this study was to develop a method to quantitatively delineate the pre-ictal window in order to prevent arbitrary pre-ictal window selection in seizure prediction models in order to improve neural modulation techniques and improve the understanding of neural dynamics that underlie seizure transition. By using a sliding window logistic regression classifier, we were able to test a spectrum of ground truths for a hypothesized pre-ictal window in a tetanus toxin (TeNT) rat model of epilepsy and in a non-human primate (NHP) penicillin (PCN) model of epilepsy. Results revealed that our classifier was able to distinguish between pre-ictal and interictal windows, which were validated by AUC values greater than 0.5 and the presence of AUC plateaus, which were characterized by consecutive pre-ictal window durations with similar receiver operating characteristics (ROC) curves and AUC values. Our tool has made it possible to quantitatively delineate the pre-ictal window on a subject-specific basis and compare the pre-ictal neural dynamics between different seizure models.
Table of Contents
Abstract [1]
Introduction [2]
Methods [6]
Results [15]
Discussion [25]
Figures [31]
Figure 1 [31]
Figure 2a [32]
Figure 2b [33]
Figure 3 [34]
Figure 4 [35]
Figure 5 [36]
Figure 6 [37]
Figure 7 [38]
Figure 8 [39]
References [40]
About this Honors Thesis
| School | |
|---|---|
| Department | |
| Degree | |
| Submission | |
| Language |
|
| Research Field | |
| Keyword | |
| Committee Chair / Thesis Advisor | |
| Committee Members |
Primary PDF
| Thumbnail | Title | Date Uploaded | Actions |
|---|---|---|---|
|
|
Pre-Ictal Window Identification via Logistic Regression () | 2020-05-02 08:40:09 -0400 |
|
Supplemental Files
| Thumbnail | Title | Date Uploaded | Actions |
|---|