Abstracts

The NeuroPace Responsive Neurostimulator System (RNS) is a fully implanted device currently being tested in a clinical trial. The RNS records electrocorticograms (ECoG) from subdural strip and depth electrodes implanted in or near the seizure focus when epileptiform discharges are detected. These ECoGs can later be retrieved and analyzed. This trial provides the first opportunity to analyze ambulatory ictal ECoG. The goal of this study is to analyze the range of intracranial ictal discharge patterns observed in an ambulatory patient population. ECoG records, each containing 30 seconds to 4 minutes of ECoG from up to 8 electrodes, were recorded by RNS devices implanted in study patients. ECoG storage was triggered manually or by pattern detection, responsive stimulation or amplifier saturation. Thus, the sampling of ECoGs is heavily biased towards ECoGs that contain ictal discharge patterns. Artifacts from telemetry dropouts and stimulation were removed. The RNS detection algorithm was configured to detect specific ictal discharge patterns including: 1) increasing amplitude, 2) amplitude attenuation, 3) 1-4 Hz sinusoidal waves or spikes, 4) 4-30 Hz theta, alpha or beta frequency patterns, and 5) 13-125 Hz gamma frequency activity. The RNS detection algorithm was applied to all of the ECoG records and ECoG classifications were visually corroborated. Descriptive statistics were calculated for each ictal pattern. Within each patient, the number of ECoGs containing each pattern was calculated and the patterns were categorized by electrode location. A preliminary analysis using a computerized [quot]increasing amplitude[quot] detector tuned to a 150% area under-the-curve increase comparing a 4-second short-term window with a 2-minute long-term window produced the following results. From 47 patients, 35,838 ECoG records totaling 634.4 hours were analyzed. Of these, 7,637 (21%) of the ECoGs contained the [quot]increasing amplitude[quot] pattern. ECoGs containing this pattern were recorded from temporal (61%), frontal (22%), parietal (11%) and other (6%) areas. For each patient, the percentage of ECoGs containing the pattern was calculated. Across all patients, the average rate of occurrence of pattern-containing ECoGs was 20.5%[plusmn]16.6% SD, ranging from 0.3% to 67.7%. Additional analysis will compare the rate of occurrence of the various ictal discharge patterns with respect to brain area and etiology. The results of this analysis suggests that chronic ambulatory electrocortigraphy patterns and frequencies are identical to those observed in acute electrocortography. Differences in ictal discharge patterns between brain areas and epilepsy etiologies are likely to be observed. (Supported by NIST Advanced Technology Program Cooperative Agreement No. 70NANB3H3044.)