Authors :
Presenting Author: Ryan Shores, BA – Rice University
Yash Vakilna, MS – University of Texas Health Science Center at Houston
Takfarinas Medani, PhD – University of Southern California
Anand Joshi, PhD – University of Southern California
Camryn Matthews, BS – University of Texas Health Science Center at Houston
Richard Leahy, PhD – University of Southern California
John Mosher, PhD – University of Texas Health Science Center at Houston
Nitin Tandon, MD – McGovern Medical School, University of Texas Health Science Center at Houston
Sandipan Pati, MD – University of Minnesota
John Seymour, PhD – University of Texas Health Science Center at Houston
Rationale:
Epilepsy diagnostic tools are tasked with locating the epileptogenic zone (EZ) – an unknown volume of tissue that varies by individual and is surrounded by both the irritative zones it creates and normal tissue. Non-invasive scalp electroencephalography (EEG) is commonly used in presurgical evaluations to approximate the EZ, but volume conduction makes signal localization difficult. Stereo-electroencephalography (sEEG) with ring electrode arrays is the current clinical standard for invasive monitoring in epilepsy patients; however, electrodes of this shape require triangulation between multiple arrays to achieve localization. Directional and scalable (DiSc) electrode arrays have been demonstrated to record local field potentials (LFPs) in rat cortex with a spatial sensitivity, amplitude and SNR that surpasses ring electrodes. Thus, significant improvement in seizure monitoring may be possible if depth arrays that can provide standalone directionality are used in place of ring arrays.
Methods:
Two 64-channel DiSc depth arrays implanted into the left hippocampus recorded intracranial EEG while surface electrodes concurrently recorded skull electrocorticography (ECoG) in Sprague Dawley rats (Figure 1). In vivo LFP recordings were taken under anesthesia before and after the left basolateral amygdala was kindled with kainic acid (KA). For comparison, copies of the DiSc microelectrode signals were spatially averaged to emulate ring electrode arrays with various contact sizes implanted in the same location (Figure 2). Template magnetic resonance imaging (MRI) segmentations from the Waxholm Space Atlas of the Sprague Dawley Rat Brain were then normalized to new, post-mortem, T2 MRI scans of each subject to contextualize the placement. The devices were mapped to the same coordinate system as the segmented MRI and the signals were compared.Results:
In referential LFP montages, the DiSc arrays’ increased spatial resolution showed complex patterns of phase reversals around the device that were not visible in ring electrodes. Consistent with the passive spatial averaging observed when using large metal contacts, the virtual rings’ larger surface area produced a muted SNR and waveforms that were less representative of the surrounding tissue. When monitoring the ratio of signal line length (LLR) after KA administration to the average of a pre-injection baseline, the DiSc arrays revealed instances of increased LLR driven by spatially distinct regions around the circumference of the device while ring electrodes could only differentiate depth. Atlas normalization allowed DiSc arrays to localize most of these sources to the ventral hippocampus despite being implanted between the hippocampus and the thalamus.Conclusions:
In conclusion, DiSc microelectrode arrays have an intrinsic amplification and directionality that can be used to create rich anatomical iEEG datasets in seizing subjects. The epileptiform events from DiSc exhibited higher amplitude, directional information, and dipole orientation relative to micro- and macroscale ring structures.Funding:
UTHealth Neurology, NIH NINDS 1RF1NS133972-01 and 1UG3NS125487-01.