Multidien Rhythm Quantification Across 4,000 Patients Implanted with the Neuropace RNS® System
Abstract number :
2.19
Submission category :
3. Neurophysiology / 3G. Computational Analysis & Modeling of EEG
Year :
2024
Submission ID :
1226
Source :
www.aesnet.org
Presentation date :
12/8/2024 12:00:00 AM
Published date :
Authors :
Presenting Author: Jacob Norman, PhD – NeuroPace, Inc.
Thomas Tcheng, PhD – NeuroPace
Sharanya Arcot Desai, PhD – NeuroPace, Inc.
Rationale: The spontaneous nature of seizures causes significant disruption for patients with epilepsy. The ability to forecast seizures would allow patients to take precautions during high-risk times. Recent work has established the relationship between seizure timing and cycles of interictal epileptiform activity (IEA). These IEA cycles were shown to be present in a majority of patients, in particular at circadian (~1 day) and multidien (4-40 days) periodicities (Baud et al., 2018; Leguia et al., 2021). In this work, we explore the presence and patterns of multidien rhythms across patient populations using the largest ambulatory intracranial EEG dataset to date.
Methods: We leveraged IEA detection data from 6,139 patients using the NeuroPace RNS® System to identify multidien rhythms. Analysis was focused on the most recent 12 months for each patient, and patients without sufficient data were excluded from analysis. Additionally, patients with less than 15 months of data were excluded to account for a 3-month implant effect (Sun et al., 2018). Detection counts for each patient were normalized between device programming changes, then wavelet decomposition was conducted to identify oscillation amplitudes for each periodicity. Multidien rhythms for each patient were identified using a peak finding algorithm, and if multiple peaks were present, the most prominent peak was selected.
Results: Of the 6,139 patients analyzed, 3,987 patients met the criteria for inclusion in this analysis. Around 55% of these patients exhibited multidien rhythms with sufficient strength. These rhythms clustered around periodicities of ~7 days, ~21 days, and ~1 month (Figure 1). Further segmentation of the patient population showed that males, on average, have slightly shorter multidien rhythms than females (p < .05, X2 test). Finally, periods associated with greater seizure reduction were less likely to contain multidien rhythms.
Conclusions: This research quantifies multidien rhythmicity of IEA in a large patient dataset, which allows for further segmentation of the patient population. Given the previously demonstrated relationship between IEA rhythmicity and seizures, these patterns in multidien rhythms could provide an avenue toward seizure forecasting.
Funding: n/a
Neurophysiology