TY - GEN
T1 - A low-cost real-time closed-loop epileptic seizure monitor and controller
AU - Young, Chung Ping
AU - Hsieh, Chao Hsien
AU - Wang, Hsu Chuan
PY - 2009/11/25
Y1 - 2009/11/25
N2 - Epilepsy is a neurological disorder, which sometimes cannot be successfully treated. We propose a real-time closed-loop monitoring and controlling device for epileptic seizure detection and suppression. This wireless-networked embedded device includes signal conditioning circuitry, a stimulator, and a microcontroller with a wireless transceiver. A TI CC2430 receives the conditioned EEG signals and performs feature extraction on them to determine if a seizure has happened. The ZigBee-based wireless transceiver transmits the EEG data to the backend computer for future off-line study. If a seizure is detected after the real-time computation, an enabling signal is sent to the stimulator to generate stimulating pulses to suppress the seizure. The feature extraction is implemented using entropy and spectrum analysis, followed by an LLS classifier. A fast seizure detection response time of around 0.6 s and a seizure detection algorithm accuracy of above 95%, when applied to a standard dataset, were achieved with the proposed portable embedded device.
AB - Epilepsy is a neurological disorder, which sometimes cannot be successfully treated. We propose a real-time closed-loop monitoring and controlling device for epileptic seizure detection and suppression. This wireless-networked embedded device includes signal conditioning circuitry, a stimulator, and a microcontroller with a wireless transceiver. A TI CC2430 receives the conditioned EEG signals and performs feature extraction on them to determine if a seizure has happened. The ZigBee-based wireless transceiver transmits the EEG data to the backend computer for future off-line study. If a seizure is detected after the real-time computation, an enabling signal is sent to the stimulator to generate stimulating pulses to suppress the seizure. The feature extraction is implemented using entropy and spectrum analysis, followed by an LLS classifier. A fast seizure detection response time of around 0.6 s and a seizure detection algorithm accuracy of above 95%, when applied to a standard dataset, were achieved with the proposed portable embedded device.
UR - http://www.scopus.com/inward/record.url?scp=70449894243&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=70449894243&partnerID=8YFLogxK
U2 - 10.1109/IMTC.2009.5168743
DO - 10.1109/IMTC.2009.5168743
M3 - Conference contribution
AN - SCOPUS:70449894243
SN - 9781424433537
T3 - 2009 IEEE Intrumentation and Measurement Technology Conference, I2MTC 2009
SP - 1768
EP - 1772
BT - 2009 IEEE Intrumentation and Measurement Technology Conference, I2MTC 2009
T2 - 2009 IEEE Intrumentation and Measurement Technology Conference, I2MTC 2009
Y2 - 5 May 2009 through 7 May 2009
ER -