Pamula Venkata Rajesh, Valero-Sarmiento Jose Manuel, Yan Long, Bozkurt Alper, Hoof Chris Van, Helleputte Nick Van, Yazicioglu Refet Firat, Verhelst Marian
IEEE Trans Biomed Circuits Syst. 2017 Jun;11(3):487-496. doi: 10.1109/TBCAS.2017.2661701. Epub 2017 May 19.
A compressive sampling (CS) photoplethysmographic (PPG) readout with embedded feature extraction to estimate heart rate (HR) directly from compressively sampled data is presented. It integrates a low-power analog front end together with a digital back end to perform feature extraction to estimate the average HR over a 4 s interval directly from compressively sampled PPG data. The application-specified integrated circuit (ASIC) supports uniform sampling mode (1x compression) as well as CS modes with compression ratios of 8x, 10x, and 30x. CS is performed through nonuniformly subsampling the PPG signal, while feature extraction is performed using least square spectral fitting through Lomb-Scargle periodogram. The ASIC consumes 172 μ W of power from a 1.2 V supply while reducing the relative LED driver power consumption by up to 30 times without significant loss of relevant information for accurate HR estimation.
本文提出了一种具有嵌入式特征提取功能的压缩采样（CS）光电容积脉搏波描记法（PPG）读出方法，可直接从压缩采样数据中估计心率（HR）。它集成了一个低功耗模拟前端和一个数字后端，以执行特征提取，从而直接从压缩采样的PPG数据中估计4秒间隔内的平均HR。该专用集成电路（ASIC）支持均匀采样模式（1倍压缩）以及压缩比为8倍、10倍和30倍的CS模式。通过对PPG信号进行非均匀欠采样来执行CS，而使用通过Lomb-Scargle周期图的最小二乘谱拟合来执行特征提取。该ASIC在1.2V电源下功耗为172μW，同时将相对LED驱动器功耗降低多达30倍，而不会在准确估计HR的相关信息方面造成重大损失。
