Abstracts

RATIONALE: To develop and implement a multi-purpose, user-friendly system using as cue the real-time output of the FHS seizure detection algorithm for inpatient monitoring, neuropsychologic testing, warning, and therapy.
METHODS: Three PCs, external hard disks, a laserjet printer, a Neoped 4000, and two Grass S-12 stimulators were integrated via a custom network, and electrically isolated from subject. System software was developed in Visual C++, MATLAB, and Visual Basic.
RESULTS: To date, over 9500 total hours of continuous ECoG recordings (from a total of 48 subjects) have been successfully collected, displayed, and analyzed in real-time. Over 900 closed-loop stimulations have been safely delivered to 7 subjects, and over 2500 automated neuropsychologic tests have been administered to 30 subjects randomly and in response to seizure detections. All seizure detections and stimulations are logged continuously, along with time and site of onset, intensity and duration. These and other quantitative measures are available for real-time review over the network via an automated reporting system. Access to past data is also available for more detailed visual and automated analysis and printing.
CONCLUSIONS: This system demonstrates the feasibility of incorporating a number of important functions and advanced capabilities not found in any commercially available or custom made system into a practical system for inpatient monitoring. Widespread use of systems of this type will greatly advance the field of epilepsy by enabling automated testing, warning, and early (closed-loop) seizure intervention.
Support: NIH SBIR 5R44NS34630-03; Medtronic, Inc.
Disclosure: Grant - Medtronic to KUMC. Consulting - FHS from Medtronic. Ownership - FHS. Materials - Medtronic to FHS. Other - FHS seizure detection algorithm licensed by FHS from KUMC and sublicensed by Medtronic, Inc. from FHS.