A real-time compressed sensing-based personal electrocardiogram monitoring system

Abstract

Abstract-Wireless body sensor networks (WBSN) hold the promise to enable next-generation patient-centric mobilecardiology systems. A WBSN-enabled electrocardiogram (ECG) monitor consists of wearable, miniaturized and wireless sensors able to measure and wirelessly report cardiac signals to a WBSN coordinator, which is responsible for reporting them to the tele-health provider. However, state-of-the-art WBSN-enabled ECG monitors still fall short of the required functionality, miniaturization and energy efficiency. Among others, energy efficiency can be significantly improved through embedded ECG compression, which reduces airtime over energy-hungry wireless links. In this paper, we propose a novel real-time energy-aware ECG monitoring system based on the emerging compressed sensing (CS) signal acquisition/compression paradigm for WBSN applications. For the first time, CS is demonstrated as an advantageous real-time and energy-efficient ECG compression technique, with a computationally light ECG encoder on the state-of-the-art Shimmer TM wearable sensor node and a realtime decoder running on an iPhone (acting as a WBSN coordinator). Interestingly, our results show an average CPU usage of less than 5% on the node, and of less than 30% on the iPhone.

On U-Statistics and Compressed Sensing I: Non-Asymptotic Average-Case Analysis

“…Observe that the term 2 · (k/n) log(n/k) is at most 1, and vanishes with high undersampling (small k/n). Hence (17) and (19) are similar from a rate standpoint.…”

“…We conclude the following: for exactly k-sparse signals the rate (19) suffices to recover at least 1 − 3u fraction of sign-subset (β β β, S) pairs. While const in (19) must be at least 4 (recall that Figure 4(c) was somewhat pessimistic), for matrices with Gaussian entries we empirically find that const is inherently smaller, whereby const ≈ 1.8.…”

“…In the more general setting for approximately k-sparse signals, we can also have rate (19). To see this, observe that Proposition 2 also delivers an exponential bound for the worst-case projections condition, see (12).…”

“…Random sampling has certain desirable simplicity/efficiency features -see [16] on data acquisition in the distributed sensor setting. Also recent hardware implementations point out energy/complexity-cost benefits of implementing pseudo-random binary sequences [17]- [19]; these sequences mimic statistical behavior. Non-asymptotic analysis is particularly valuable, when random samples are costly to acquire.…”

A quality-scalable and energy-efficient approach for spectral analysis of heart rate variability

“…Interestingly, ECG signals are known to be approximately sparse in the wavelet domain which was exploited for compressing them [12,18], motivating us to investigate if such an attribute applies also to the RR-intervals (inputs to PSA).…”