| Epilepsy is a neurological disease.When a generalized tonic-clonic seizures occurs,the patient’s skeletal muscles contract and clonic,which easily leads to various accidents.At present,the video electroencephalograph for epilepsy monitoring has the problems of large size and harsh collection environment,which is not convenient to wear in daily monitoring.Therefore,a wearable generalized tonic-clonic seizures monitoring system is developed to monitor the patient’s life status in real time to avoid accidents during epileptic seizures,which is of great significance to improve the quality of life of patients.The wearable generalized tonic-clonic seizures monitoring system in this article is composed of the front-end signal acquisition circuit and the monitoring mobile phone software.It can send an alarm when the seizure occurs.Firstly,according to the characteristics of the changes in the physiological parameters of patients with generalized tonic-clonic seizures,we select appropriate physiological parameters to monitor the seizures Secondly,through the study of wearability,hardware low-power performance and low-power Bluetooth architecture,we designed a wearable,low-power main control circuit and physiological parameter signal acquisition circuit,including electrodermal activity signal,heart rate signal,triaxial acceleration signal and skin temperature signal acquisition circuit.Then,by studying the performance characteristics of the PCB antenna and ceramic antenna,the ceramic antenna is selected as the transmitting antenna of the monitoring system,and the vector network analyzer is used to determine the π-type matching network parameters of the antenna,the antenna performance is optimized.The minimum return loss is about-27.69 dB,and the bandwidth at-10dB is about 100MHz.Next,by optimizing the circuit of the main control chip,dynamically adjusting the transmission power,and using the low power consumption mode of the peripheral monitoring chip,the system power consumption is further optimized.Finally,according to the characteristics of the monitoring signal,the type of noise,and the real-time processing capability and the complexity of the algorithm of the embedded system,the moving average filtering algorithm is used to preprocess the monitoring signal,and the sliding window integral algorithm is used to process the acceleration signal vector magnitude in order to obtain the total amount of limb movement in a short time.According to the changes of various physiological parameter thresholds during generalized tonic-clonic seizures,an epilepsy monitoring algorithm based on multi-thresholds is proposed,and the algorithm is embedded in the hardware system.Tests show that the system is stable.The size of the monitoring watch is 4cm*4cm*1cm,and the monitoring watch uses a 270mAh lithium battery to work continuously for more than 24 hours,which can meet the needs of daily wearable monitoring.The algorithm test of the clinical monitoring data of 15 epilepsy patients is performed,then the test result is compared with the monitoring result of the video electroencephalograph,and the accuracy rate of seizure recognition is 91%. |
