TL;DR
This paper introduces Epios, a novel full-head subscalp EEG system that can be implanted with minimally invasive surgery. The study demonstrates its safety and feasibility, showing that Epios is non-inferior to conventional scalp EEG for capturing brain oscillations and pathological discharges, paving the way for long-term, unobtrusive brain monitoring.
What Problem Does This Paper Address?
Current clinical neurophysiology relies heavily on conventional electroencephalography (EEG), which typically involves short recordings (days) using electrodes placed on the scalp. While effective for capturing transient brain activity, this approach has limitations for monitoring chronic brain disorders like epilepsy over extended periods. The instability of scalp electrodes, the need for frequent maintenance, and the impracticality and stigmatizing nature of wearing EEG equipment continuously hinder long-term, unobtrusive monitoring. Intracranial EEG systems exist but are invasive and offer limited recording capabilities. This paper addresses the need for a more robust, long-term, and less invasive method for comprehensive brain monitoring, particularly for conditions requiring continuous assessment.
Methods
The study developed and tested the Epios system, which utilizes a minimally invasive technique to implant subscalp electrodes. The core innovation is a "trident" electrode lead, a flexible, silicon-based lead with platinum-iridium contacts, designed to be tunneled beneath the scalp and over the cranium.
Key methodological aspects include:
- Minimally Invasive Implantation: The trident leads were inserted through small (≤1 cm) scalp incisions using custom-made tunneling tools. This procedure was performed concurrently with clinically indicated intracranial EEG implantation in participants with pharmacoresistant epilepsy.
- Full-Head Coverage: The system aims for full-head coverage, with up to 28 electrode contacts distributed across the entire scalp area.
- Safety Monitoring: Postoperative safety was assessed through daily examinations for adverse events like scalp edema, hematoma, and pain, as well as serial pain level assessments.
- Performance Evaluation: The performance of subscalp EEG was compared to conventional scalp EEG and, where applicable, intracranial EEG (considered the ground truth for epileptiform discharges). This involved:
- Visual Sleep-Wake Scoring: Trained scorers evaluated sleep-wake architecture from both subscalp and scalp EEG recordings.
- Evoked Potentials: The ability to measure evoked potentials from visual and auditory stimuli was assessed.
- Spectral Analysis: Power spectral density and coherence were analyzed to compare brain oscillations.
- Topographical Analysis: Topographic maps were generated to visualize brain activity distribution and assess spatial correlation between subscalp and scalp EEG.
- Ictal and Interictal Discharge Analysis: The system's ability to record and localize epileptic seizures (ictal) and interictal spikes was evaluated against intracranial EEG.
- Non-inferiority Testing: Statistical methods, including Bland-Altman analysis and calculation of interclass correlation coefficients (ICC), were used to determine if subscalp EEG was non-inferior to scalp EEG for various measures.
Datasets
The study utilized data from 8 adult participants with pharmacoresistant epilepsy who were undergoing intracranial EEG monitoring for presurgical seizure localization. These participants received the investigational subscalp electrodes in addition to their clinical intracranial and scalp EEG setups. No new benchmark datasets were introduced; rather, the study used existing clinical EEG recordings as a comparator.
Key Insights and Ideas
- Feasibility of Full-Head Subscalp Coverage: The most significant insight is that it's possible to safely implant a substantial number of subscalp electrodes (up to 28) across the entire head using a minimally invasive surgical approach. This overcomes a major hurdle for long-term, comprehensive brain monitoring.
- Safety Profile: The procedure demonstrated a favorable safety profile, with only minor, self-limited adverse events observed, and no serious device-related complications. This suggests the technique is well-tolerated.
- Non-Inferiority to Scalp EEG: A crucial finding is that subscalp EEG is not inferior to conventional scalp EEG in capturing a wide range of brain activity, including physiological oscillations (e.g., alpha, delta, sigma, beta) during different sleep-wake states and pathological discharges like interictal spikes and ictal events.
- Robustness to Artifacts: The subscalp system showed resilience to artifacts during various activities, suggesting its potential for reliable recording in daily life.
- Localization Capability: The full-head coverage of the subscalp electrodes allows for the localization of brain activity with lobar or sublobar precision, which is critical for understanding and managing neurological disorders like epilepsy.
Key Results
- Primary Outcome (Feasibility & Safety): Up to 28 subscalp electrodes were safely placed over the entire head through 1-cm scalp incisions in approximately a 1-hour procedure. Five minor perioperative adverse events were linked to the procedure, but none were serious and all resolved.
- Secondary Outcome (Performance):
- Wake EEG: Subscalp EEG was non-inferior to scalp EEG in measuring posterior-dominant peak alpha power (ICC = 0.86, bias = 0.03 dB). Spatial correlation between subscalp and scalp topoplots was moderate to strong (p = 0.74).
- Evoked Potentials: Subscalp EEG showed borderline agreement for visual evoked potentials (ICC = 0.26) and strong agreement for auditory evoked potentials (ICC = 0.80), with bias within the non-inferiority margin.
- Sleep EEG: Subscalp EEG was non-inferior to scalp EEG for measuring N2 sigma, N3 delta, and REM beta oscillations (ICCs ranging from 0.86 to 0.96). Spatial correlations for sleep oscillations were moderate to strong. Sleep scoring agreement between subscalp and scalp EEG was high (Cohen's Kappa > 0.67).
- Ictal EEG: Subscalp EEG was able to capture a high percentage of seizures (78%) and showed strong coherence with intracranial EEG in the seizure onset zone (SOZ) and propagation zone (PZ) (ICCs up to 0.92).
- Interictal EEG: Subscalp EEG was non-inferior to scalp EEG in measuring and localizing interictal spike peak amplitudes (ICC = 0.81) and showed strong spatial correlation (p = 0.92).
Limitations and Future Work
- Short Duration: The primary limitation is the short duration of the trial (up to 9 days) and the limited number of participants. This restricts definitive conclusions about long-term safety and utility.
- Selection Bias: The study acknowledges a selection bias, as only 57% of screened patients agreed to participate, potentially favoring those with more severe epilepsies or deeper seizure foci.
- Hospital Setting: The trial was conducted in a hospital setting, not in an ambulatory environment.
Future Work:
- Longer Prospective Ambulatory Study: The authors emphasize the need for a longer prospective ambulatory study to establish the long-term safety and utility of the full Epios system.
- Ambulatory System Development: The current study focused on the implantable lead. Future work will involve connecting these leads to an implanted transceiver for transcutaneous data collection, enabling truly ambulatory monitoring over many months.
- Broader Applications: The potential for this technology extends beyond epilepsy to other brain disorders requiring continuous monitoring and localization of brain activity.
Summary updated