摘要
提出一种基于SoC并利用PVDF非接触式传感带对多维度生理信号进行实时处理的硬件化算法。从PVDF压电薄膜传感带中获取生理信号,经过前端电荷放大、滤波、ADC采样后进入可编程逻辑器件FPGA。算法采用流水线设计,提取生理信号的特征值、计算体动合成指数,通过软核处理器Nios接收硬件化计算结果,对睡眠状态进行最终识别。经实验对比验证,该算法能在19 ms内识别人体睡眠时的肢体律动、正常呼吸、呼吸暂停3种状态,极大地缩短了软件处理的时间,同时为医生诊断呼吸暂停综合症提供病理依据,在医院临床和家庭监护方面具有参考价值。
This paper introduces an algorithm using unobtrusive PVDF sensor to acquire and process multi-dimension physiological signals in real time based on So C.The system aquires the physiological signals from the PVDF piezoelectric film sensor,the signals step into the programmable logic device after the front-end charge amplification,filtering and ADC sampling.The algorithm uses pipelined logic design to extract BCG physiological signal characteristic value and generate the stabile measurement of BHI,receives the results of hardware calculation through the soft-core processor Nios,and presents the final discrimination of sleep state.The ex-perimental results show that the proposed algorithm could identify three types of sleep state,including body movements,normal res-piration and apnea events.Processing time is sharply reduced,meanwhile,the algorithm provides great reference value in medical dignose and home health care for apnea hypopnea syndrome.
出处
《电子技术应用》
北大核心
2017年第8期62-65,共4页
Application of Electronic Technique
