Development of implantable devices for automated detection, quantification, warning and delivery of therapy to block seizures is a very important unmet medical need. Making such a device as small as possible, minimizing replacement surgeries, and maximizing device longevity and/or time between battery recharging are some of the most important development drivers in a "patient-centric" design and are closely related to commercial viability of the device product. While the benefits of endowing such devices with intelligence (i.e., early warning capabilities and means for objectively quantifying seizures with high accuracy) is clear, the severe power consumption and processor speed limitations associated with the digital microprocessors used in today's implantable devices are a significant hurdle in implementing even the most efficient digital algorithms. Development of analog algorithms for use in existing and future devices provides a viable and effective way to overcome this hurdle. The focus of this proposal is on validating an ultra-low-power analog seizure detection algorithm (ASDA), which is the world's first to be implemented, completely in analog. In a small-scale preliminary study, the ASDA's performance was equivalent to that of an existing, rigorously and successfully validated, state-of-the-art digital detection algorithm. Moreover, it is estimated that the ASDA can achieve this level of performance while consuming 25-50 times less power. The ASDA's performance will be evaluated on a previously collected and visually scored multicenter database of brain signals from 130 subjects containing several thousand seizures. A detailed comparison of results will be made with the digital Osorio-Frei SDA. The existing breadboard analog implementation will also be ported to a printed circuit board version. The resulting analog seizure detection system is expected to markedly increase longevity and commercial viability of implanted devices for real- time detection, warning, and seizure blockage, while retaining superior accuracy.
Public Health Relevance: The focus of this Phase I SBIR proposal will be on validating the use of a novel signal processing technology to enable full implementation and eventual commercialization of the world's first ultra-low power completely analog seizure detection algorithm. The validation will compare the new, ultra-low-power approach with an already rigorously and successfully validated, state-of-the-art digital seizure detection algorithm. The analog algorithm's performance will be evaluated on a previously collected and visually scored database of brain signals from 130 subjects containing several thousand seizures. Device power consumption, compared to that using a conventional digital implementation, is expected to be decreased by a factor of approximately 25-50. The resulting analog seizure detection algorithm is expected to markedly increase longevity and commercial viability of implanted devices for real-time seizure detection, warning, and blockage.
Public Health Relevance: This Public Health Relevance is not available.
Thesaurus Terms: There Are No Thesaurus Terms On File For This Project.
