Amplitude demodulation of the electrocardiogram signal (ECG) for respiration monitoring and compensation during MR examinations

Abstract

A method for respiration monitoring during MR-sequences using the electrocardiogram signal (ECG) is presented. The basic principle is known in conventional patient monitoring and is based on the fact that the amplitude of the ECG-signal varies with respiratory motion. A demodulation of this "cardio-respiratory signal" yields a respiration curve that can be used for patient monitoring, triggering, or respiration compensation of MRI sequences. Since no drift is introduced by digital demodulation, this method is superior to analog signal processing. The proposed method was applied during MR examinations using an optically coupled ECG sensor that has the advantage of almost negligible interference due to the MR sequences (< < 10% of the main ECG signal). The respiration signal obtained is strongly correlated (r = 0.79 to 0.98) with the position of the diaphragm in inferior-superior direction measured during breath-holding MRI.

Novel real-time R-wave detection algorithm based on the vectorcardiogram for accurate gated magnetic resonance acquisitions

“…However, the frequency difference of this noise and the ECG spectral components itself is in the magnitude of 7 decades, and thus can be removed by readily available standard signal filtering approaches (28)(29)(30). Further, baseline drift due to respiratory motion is amplified in a static magnetic field, but can also be suppressed by high-pass filtering (31). A second criteria for the development of useful QRS detection algorithms is their feasibility for real-time detection.…”

In vivo 1H-MR spectroscopy of the human heart

“…For method C (Fig. 2), we used the fact that the amplitude of the ECG depends on respiration and the rotation of the heart [8], i.e. the position of the heart can be monitored with the ECG signal alone with an adequate positioning of the ECG electrodes.…”

“…Since the motion of the diaphragm is smallest during expiration [8], all respiration triggering was done in expiration. ECG gating alone was not adequate to stabilize the MR signal sufficiently to allow for satisfactory shimming, water suppression and spectra acquisition.…”

Adaptive RR prediction for cardiac MRI

“…The signal of the belt is non-uniformly sampled, so we used only the samples when HR is computed. As described in [6], the magnitude of QRS is correlated with respiration, this information should be very useful during breath-holds, since it is always available. The assumption done is that the RR interval is a linear combination of the previous RR, the respiration, its derivation and the magnitude of the previous RR.…”