The design and hardware implementation of a low-power real-time seizure detection algorithm

Abstract

Epilepsy affects more than 1% of the world's population. Responsive neurostimulation is emerging as an alternative therapy for the 30% of the epileptic patient population that does not benefit from pharmacological treatment. Efficient seizure detection algorithms will enable closed-loop epilepsy prostheses by stimulating the epileptogenic focus within an early onset window. Critically, this is expected to reduce neuronal desensitization over time and lead to longer-term device efficacy. This work presents a novel event-based seizure detection algorithm along with a low-power digital circuit implementation. Hippocampal depth-electrode recordings from six kainate-treated rats are used to validate the algorithm and hardware performance in this preliminary study. The design process illustrates crucial trade-offs in translating mathematical models into hardware implementations and validates statistical optimizations made with empirical data analyses on results obtained using a real-time functioning hardware prototype. Using quantitatively predicted thresholds from the depth-electrode recordings, the auto-updating algorithm performs with an average sensitivity and selectivity of 95.3 +/- 0.02% and 88.9 +/- 0.01% (mean +/- SE(alpha = 0.05)), respectively, on untrained data with a detection delay of 8.5 s [5.97, 11.04] from electrographic onset. The hardware implementation is shown feasible using CMOS circuits consuming under 350 nW of power from a 250 mV supply voltage from simulations on the MIT 180 nm SOI process.

An algorithm for real-time detection of spike-wave discharges in rodents

“…However, all of them lack the ability to process real-time data. While some reports have described realtime SWD detection (Raghunathan et al, 2009;Talathi et al, 2008), the proposed techniques have all shown relatively low performance. Here we propose and evaluate a CWT method for real-time detection of SWDs in ECoG using rats of the WAG/Rij strain (van Luijtelaar and Coenen, 1986;van Luijtelaar and Sitnikova, 2006), and compare its performance against visual scoring.…”

A hardware-algorithm co-design approach to optimize seizure detection algorithms for implantable applications

“…The marked out seizures included both clinical and sub-clinical events based on electrographic inspection. We have previously documented these methods in (Raghunathan et al, 2009a). The seizures were classified from 1 to 5 on a Racine scale, per published protocols (Hellier and Dudek, 2005).…”

“…The Neuropace ® device provides a platform for incorporating highly efficient, hardware-optimized detection algorithms that can be implemented using microchips that can operate with orders of magnitude lower power consumption than standard microcontrollers (Raghunathan et al, 2009a). The analysis performed in this paper compares a set of published seizure detection features and evaluates them from a dual viewpoint of both detection and hardware efficiency.…”

Seizure detection methods using a cascade architecture for real-time implantable devices

“…Most notably, recent studies [5][6][7][8] ing low power on real-time implantable devices. However, these low power, computationally-efficient (i.e., lightweight) seizure detection algorithms use only timedomain features including indirect frequency features such as count-based, and inter-event due to the computational complexity of frequency-domain feature computation.…”