Authors :
Presenting Author: Jeffrey Tenney, MD, PhD – Cincinnati Children's Hospital Medical Center
Seungrok Hong, B.S. – School of Medicine – University of Cincinnati; Paul Horn, PhD – Cincinnati Children's Hospital Medical Center; Hisako Fujiwara, PhD – Cincinnati Children's Hospital Medical Center; Jesse Skoch, MD – Cincinnati Children's Hospital Medical Center; Ravindra Arya, MD – Cincinnati Children's Hospital Medical Center; Gewalin Aungaroon, MD – Cincinnati Children's Hospital Medical Center; Susan Fong, MD, PhD – Cincinnati Children's Hospital Medical Center; Katherine Holland-Bouley, MD, PhD – Cincinnati Children's Hospital Medical Center; Kelly Kremer, MD – Cincinnati Children's Hospital Medical Center; Katie Ihnen, MD, PhD – Cincinnati Children's Hospital Medical Center; Nan Lin, MD – Cincinnati Children's Hospital Medical Center; Wei Liu, MD – Cincinnati Children's Hospital Medical Center; Heather Wied, MD, PhD – Cincinnati Children's Hospital Medical Center; Francesco Mangano, DO – Cincinnati Children's Hospital Medical Center; Hansel Greiner, MD – Cincinnati Children's Hospital Medical Center
Rationale:
The most common medically resistant epilepsy (MRE) involves the temporal lobe (TLE) and 25% of patients continue to suffer from seizures after temporal lobectomy, with children designated as temporal plus epilepsy (TLE+) having a 5 times increased risk of post-operative surgical failure. Magnetoencephalography (MEG) is used routinely to provide targets for subsequent intracranial EEG (iEEG). The most common, well validated MEG source localization algorithm is the equivalent current dipole (ECD) however it is often difficult to summarize the richness of MEG data with one or a few point sources. The objective was to correlate the visual analysis of MEG virtual sensor waveforms with TLE and TLE+. The hypothesis is that user-defined virtual sensor beamforming (UDvs-beamforming), using expert reader analysis, is superior to ECD, current density source modeling, and conventional beamforming for non-invasively differentiating between TLE and TLE+.
Methods:
Patients with MRE who (1) underwent MEG, (2) iEEG monitoring, and (3) have at least one year post-surgical follow-up were included in this retrospective analysis. Participants with only temporal foci determined by iEEG were grouped as TLE and those with temporal and extra-temporal foci were deemed TLE+. Conventional MEG analysis included ECD, sLORETA, and synthetic aperture magnetometry (SAM) beamformer analyses. UDvs-beamforming was completed with virtual sensors placed manually and symmetrically in the bilateral amygdalohippocampi, inferior/middle/superior temporal gyri, insula, suprasylvian operculum, orbitofrontal cortex, and TPO junction at 10mm spacing. Each of the 12 locations was classified as positive or negative for ECD and sLORETA dipoles, SAM beamformer activity, and UDvs-beamformer spikes. A logistic regression was used to examine the interaction of these MEG findings to both patient group (TLE/TLE+) and surgical outcome (seizure free (ILAE 1), not seizure free (ILAE 2-6)).
Results:
Eighty patients (38 females, 42 males) with MRE (mean age 11.3 ± 6.2 yrs, range 1.0-31.5) were identified and included. Mean duration of epilepsy was 6.1 ± 5.8 yrs (range 0.1-28.9) with 51% lesional, 29% multi-lesional, and 20% non-lesional. Twenty-five patients (31.3%) were classified as TLE while 55 (68.8%) were TLE+. The odds of a non-seizure free outcome increased with each additional abnormal location identified with UDvs-beamforming (OR 1.28, 95% CI [1.04-1.58], p-value = 0.022). When modeling the probability of TLE+ classification, UDvs-beamforming (OR 1.47, 95% CI [1.13-1.92], p-value = 0.004) was superior to ECD (OR 1.01, 95% CI [0.67-1.51], sLORETA (OR 0.92, 95% CI [0.58-1.47]), and SAM beamformer (OR 1.04, 95% [0.64-1.67]).
Conclusions:
Our modified beamforming method (UDvs-beamforming) was correlated with seizure outcome so that each additional abnormal location increased the odds of a non-seizure free outcome by 28%. Additionally, UDvs-beamforming was correlated with TLE classification whereas conventional MEG source localization was not close to statistical significance. These findings support the use of UDvs-beamforming to non-invasively differentiate between TLE/TLE+ and more accurately predict seizure outcome.
Funding:
5R21NS123630