Simultaneous measurement of diaphragm activity, chest impedance, and ECG using three standard cardiorespiratory monitoring electrodes

Abstract

Introduction: Current cardiorespiratory monitoring in neonates with electrocardiogram (ECG) and chest impedance (CI) has limitations. Adding transcutaneous electromyography of the diaphragm (dEMG) may improve respiratory monitoring, but requires additional hardware. We aimed to determine the feasibility of measuring dEMG and ECG/CI simultaneously using the standard ECG/CI hardware, with its three electrodes repositioned to dEMG electrode locations.

Methods: Thirty infants (median postmenstrual age 30.4 weeks) were included. First, we assessed the feasibility of extracting dEMG from the ECG-signal. If successful, the agreement between dEMG-based respiratory rate (RR), using three different ECG-leads, and a respiratory reference signal was assessed using the Bland-Altman analysis and the intraclass correlation coefficient (ICC). Furthermore, we studied the agreement between CI-based RR and the reference signal with the electrodes placed at the standard and dEMG position. Finally, we explored the quality of the ECG-signal at the different electrode positions.

Results: In 15 infants, feasibility of measuring dEMG with the monitoring electrodes was confirmed. In the next 15 infants, comparing dEMG-based RR to the reference signal resulted in a mean difference and limits of agreement for ECG-lead I, II and III of 4.2 [-8.2 to 16.6], 4.3 [-10.7 to 19.3] and 5.0 [-14.2 to 24.2] breaths/min, respectively. ICC analysis showed a moderate agreement for all ECG-leads. CI-based RR agreement was similar at the standard and dEMG electrode position. An exploratory analysis suggested similar quality of the ECG-signal at both electrode positions.

Conclusion: Measuring dEMG using the ECG/CI hardware with its electrodes on the diaphragm is feasible, leaving ECG/CI monitoring unaffected.

Keywords: neonates; respiration; transcutaneous electromyography.

Figure 1

The standard electrode positions to measure the electrocardiogram (ECG) and chest impedance (CI) with the corresponding three ECG‐leads (A). The ECG/CI electrodes placed at the positions of a transcutaneous electromyography measurement of the diaphragm (dEMG), with the corresponding ECG‐leads (B).

Figure 2

Steps to obtain a respiratory waveform out of a representative ECG‐tracing with incorporated electrical activity of the diaphragm. (A) The ECG‐signal (in this case lead 2) after high pass filtering to remove the offset and drift. (B) The shifted (due to delay caused by QRS‐detection) ECG‐signal with gates around the detected QRS‐complexes (solid line) and P‐waves (dotted line, at a fixed distance). (C) The result after gating, filling with a copy of previous data and 50 Hz notch filtering. (D) The respiratory waveform after rectification and moving average filtering. ECG, electrocardiogram.

Figure 3

A representative tracing of simultaneously measured airway pressure, chest impedance (CI) and diaphragm activity. The latter was measured with transcutaneous electromyography (dEMG) incorporated in the electrocardiogram (ECG), in this case ECG‐lead I. Airway pressure, CI and dEMG are comparable in terms of the recorded breathing cycles.

Figure 4

Random 5 second tracings of the ECG‐signal measured per lead with the electrodes placed at the height of the diaphragm (top three subplots) and at the standard location (bottom, in this case lead II). Note: the four ECG‐tracings are not measured simultaneously.