T101. Use of a quantitative algorithm to help predict seizure lateralization in a patient with bitemporal epilepsy and responsive nerve stimulation. (May 2018)

• Record Type: Journal Article
• Authors: Haagensen, Jennifer J.; Chen, Stephanie; Hopp, Jennifer L.; Li, Adam; Sarma, Sridevi
• Is Part Of: Clinical neurophysiology. Volume 129(2018)Supplement 1
• Page Start: e41
• Publication Date: 2018-05
• Subjects: Neurophysiology -- Periodicals; Electroencephalography -- Periodicals; Electromyography -- Periodicals; Neurology -- Periodicals; 612.8
• DOI: 10.1016/j.clinph.2018.04.102
• Languages: English
• ISSNs: 1388-2457

Abstract

Abstract : Introduction: Bilateral mesial temporal epilepsy (Bi-mTLE) can be seen in up to 39% of patients with mTLE (Aghakhani et al., 2014). Up to 77% of Bi-mTLE patients ultimately undergo surgery, yet only 25% of these resected patients obtain an outcome of Engel Class I (Aghakhani Y et al., 2014). A network-based algorithm, EZTrack, has been developed to quantitatively illustrate the most "fragile" nodes in a network using electrocorticography (ECoG) data and may help with locating the epileptogenic zone (EZ) (Li et al., 2017). Fragility of a node is defined as the smallest amount of change to that node's influence on its neighbors required to cause a seizure and can be quantified using network models (Li et al., 2017). We present a patient with bi-mTLE who underwent Phase II intracranial monitoring and EZTrack analysis. Methods: EZTrack is an algorithm designed with the idea that the brain is a dynamic networked system. Seizures can be considered a divergence from a stable network dynamic to an unstable network dynamic. The EZTrack algorithm uses raw ECoG data and applies network analysis techniques with each ECoG electrode representing a node in a network. The ECoG data is processed to form a sequence of matrices and the fragility of each node is tracked over time. The fragility matrix is converted into a heatmap that represents the nodes that are computed as the most epileptogenic as previously described (Li et al., 2017). ECoG data from the RNS was collected from two (one right and one left) 4-contact subtemporal strips and uploaded to the patient data management system® (PDMS®) over the last 8 months, as per clinical protocol. A board-certified epileptologist reviewed ECoG to determine seizure laterality (King-Stephens et al., 2015). Results: A 54 year-old left-handed man with 20+ years of medically refractory epilepsy (MRE) underwent intracranial phase II monitoring with bilateral subtemporal strips (anterior, mesial and posterior). Four seizures were captured: no EEG correlate (#1), left temporal onset (#2 and 4), and right temporal onset (#3). EZTrack accurately calculated the regions that the epileptologist identified as the EZ for all three electroclinical seizures. Interestingly, the electrographic seizure with right temporal onset demonstrated increased fragility in the left anterior temporal 4 electrode approximately 12 s prior to electrographic onset on the right. Since RNS implantation 8 months ago, 13 electrographic seizures have been captured and recorded: 11 with left temporal onset and 2 with right temporal onset. Conclusion: EZTrack may add useful predictive information regarding seizure laterality in patients with bi-mTLE. This information could be helpful to better plan palliative resection.