A Methodology for Embedded Classification of Heartbeats Using Random Projections

Abstract

Abstract-Smart Wireless Body Sensor Nodes (WBSNs) are a novel class of unobtrusive, battery-powered devices allowing the continuous monitoring and real-time interpretation of a subject's bio-signals. One of its most relevant applications is the acquisition and analysis of Electrocardiograms (ECGs). These low-power WBSN designs, while able to perform advanced signal processing to extract information on hearth conditions of subjects, are usually constrained in terms of computational power and transmission bandwidth. It is therefore beneficial to identify in the early stages of analysis which parts of an ECG acquisition are critical and activate only in these cases detailed (and computationally intensive) diagnosis algorithms. In this paper, we introduce and study the performance of a real-time optimized neuro-fuzzy classifier based on random projections, which is able to discern normal and pathological heartbeats on an embedded WBSN. Moreover, it exposes high confidence and low computational and memory requirements. Indeed, by focusing on abnormal heartbeats morphologies, we proved that a WBSN system can effectively enhance its efficiency, obtaining energy savings of as much as 63% in the signal processing stage and 68% in the subsequent wireless transmission when the proposed classifier is employed.

A Synchronization-Based Hybrid-Memory Multi-Core Architecture for Energy-Efficient Biomedical Signal Processing

“…This benchmark, detailed in [5], embeds a neuro-fuzzy classifier that detects abnormal heartbeats. When a detection occurs, a further analysis is executed on the abnormal heartbeat.…”

“…Latest WBSNs are able to perform complex on-node Digital Signal Processing (DSP) routines, such as Electrocardiogram (ECG) compression [3], automated feature extraction [4] and classification [5]. DSP applications embedded in such "smart" WBSNs greatly reduce the required transmission bandwidth, thus increasing the overall energy efficiency.…”

“…Our approach is particularly beneficial for bio-signal DSP, where applications usually acquire multiple signals at rather low sample rates (e.g., hundreds of Hz in the case of ECGs [4] [3] [5]). The majority of run-time execution is therefore spent on acquisition alone, as opposed to data processing.…”

Hardware/software approach for code synchronization in low-power multi-core sensor nodes

“…The third benchmark (RP-CLASS) uses a heartbeat classifier, operating on a single lead, to discern normal from pathological heartbeats, applying the method proposed by the authors of [22]. When an abnormal situation is detected, a three-lead delineation is activated only for the pathological heartbeat.…”

Inexact‐aware architecture design for ultra‐low power bio‐signal analysis

“…Recently, however, a new generation of smart WBSNs has emerged which are able to perform digital signal processing directly on-board to analyse the acquired bio-signals and extract clinically-relevant features, in addition to data acquisition and transmission [4]. These devices pave the way for truly autonomous and versatile health monitoring devices.…”