arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2203.04396eess.SPcs.HCcs.LG
收藏

嵌入时间卷积网络用于能量有效的基于PPG的心率监测

Embedding Temporal Convolutional Networks for Energy-Efficient PPG-Based Heart Rate Monitoring

University of Bologna(博洛尼亚大学);Politecnico di Torino(都灵理工大学);ETH Zurich(苏黎世联邦理工学院)

Alessio Burrello, Daniele Jahier Pagliari, Pierangelo Maria Rapa, Matilde Semilia, Matteo Risso, Tommaso Polonelli, Massimo Poncino, Luca Benini, Simone Benatti

摘要:

光电体积描记(PPG)传感器允许进行非侵入性和舒适的心率(HR)监测,适用于紧凑型腕戴式设备。不幸的是,运动伪影(MA)严重影响监测精度,导致皮肤与传感器接口的高度可变性。基于将PPG信号与惯性传感器数据相结合,已经引入了几种数据融合技术来应对这个问题。在此之前,商业和重新配置的解决方案都是计算效率高的,但不是很鲁棒,或者强烈依赖于手动调整的参数,这导致泛化性能差。在这项工作中,我们通过提出一种基于PPG的HR估计的计算轻量级但强大的基于深度学习的方法来解决这些限制。具体来说,我们利用神经架构搜索(NAS)推导出一组不同的时间卷积网络(TCN)用于HR估计。此外,我们还引入了ActPPG,一种自适应算法,根据MA的量在多个HR估计器中进行选择,以提高能源效率。我们在两个基准数据集上验证了我们的方法,在PPGDalia上实现了低至3.84次/分钟(BPM)的平均绝对误差(MAE),优于以前的最先进水平。此外,我们将我们的模型部署在低功耗商用微控制器(STM32 L4)上,在复杂度与准确度空间中获得了丰富的帕累托最优解决方案。

英文摘要:

Photoplethysmography (PPG) sensors allow for non-invasive and comfortable heart-rate (HR) monitoring, suitable for compact wrist-worn devices. Unfortunately, Motion Artifacts (MAs) severely impact the monitoring accuracy, causing high variability in the skin-to-sensor interface. Several data fusion techniques have been introduced to cope with this problem, based on combining PPG signals with inertial sensor data. Until know, both commercial and reasearch solutions are computationally efficient but not very robust, or strongly dependent on hand-tuned parameters, which leads to poor generalization performance. % In this work, we tackle these limitations by proposing a computationally lightweight yet robust deep learning-based approach for PPG-based HR estimation. Specifically, we derive a diverse set of Temporal Convolutional Networks (TCN) for HR estimation, leveraging Neural Architecture Search (NAS). Moreover, we also introduce ActPPG, an adaptive algorithm that selects among multiple HR estimators depending on the amount of MAs, to improve energy efficiency. We validate our approaches on two benchmark datasets, achieving as low as 3.84 Beats per Minute (BPM) of Mean Absolute Error (MAE) on PPGDalia, which outperforms the previous state-of-the-art. Moreover, we deploy our models on a low-power commercial microcontroller (STM32L4), obtaining a rich set of Pareto optimal solutions in the complexity vs. accuracy space.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑