Brain Products Press Release
April 2013, Volume 46
IN THE FOCUS
The Respiration Belt MR: a new device for parallel respiratory measurements
by Nicola Soldati
The investigation of physiological signals continues to receive a
great amount of attention. The obvious reason is that such signals
are deeply rooted in the nature of the subject of investigation
(i.e. living beings) and they strongly interact with the organisms
at various levels, from pure physiology to higher level cognitive
functions.
Depending on the type of research, these signals can be useful
when addressing specific questions, or they maybe simply add
noise to the signals of interest. The latter is often the case in
neuroscience, where strong physiological phenomena such as
cardiac and respiratory cycles affect the measurements acquired
by different techniques (EEG, fMRI).
Respiration plays a critical role in the MR environment, where it
may not only be a confounding factor, but also a source of
related artifacts. It can be linked to movement artifacts (due to
the mechanical action of breathing - the typical respiratory rate
of a healthy adult is 12-20 breaths per minute), physiological
alterations (change of BOLD signal properties), induced field
inhomogeneity (change of air volume in the lungs can affect the
magnetic field locally), or interference with the experimental
paradigm.
Studies using fMRI show that respiratory effects cannot be
ignored, given that respiration induces great changes in terms
of artifacts, and different respiratory patterns cause different
oxygenation and finally change the fMRI measured BOLD signal
(Thomason et al. 2005).
For this reason, advanced signal processing techniques have
been developed with the goal of eliminating these confounding
factors. One proposal was the use of Independent Component
Analysis (ICA) to correct and remove structured noise (Thomas
et al. 2002). However, recent work has shown that ICA alone
cannot completely remove physiological noise from fMRI data
(Beall et al. 2010) and moreover that higher order fluctuations
in respiratory patterns induce detectable signal changes which
can act as a confounding factor in research related to resting state
(Birn et al., 2008).
Even if advances in data analysis techniques can provide better
results at the cost of greater complexity, these results are
considerably improved by parallel dedicated measurements of
the sources of the artifacts. An efficient method which exploits
parallel measurements for artifact correction uses acquired
respiratory signals to create a principal regressor, along with
other derived regressors obtained with a higher order analysis
of the signal itself. This approach is known as RETROICOR
(Glover et al., 2000). It is clear that a higher quality and sensitivity
of acquired respiratory data will lead to an improved quality
of all the regressors and finally to a higher quality of artifact
correction and final denoised data, independent of the strategy
adopted to correct for respiratory artifacts. With the aim of
obtaining the best data quality and the optimal method of
artifact correction we have developed the Respiration Belt MR,
a novel device for the acquisition of respiratory signals within
MR environments (Fig.1).
Working in an MR environment imposes several constraints
ranging from the safety and care of the subject to the quality of
the acquired data. Our solution offers advantages for all these
factors. We decided to realize a respiratory belt, because this is
a non-intrusive sensor which is comfortable for the test subjects,
who may already be negatively affected by the fMRI procedure
(Cook et al., 2007).
The compatibility and safety of the Respiration Belt MR result
from its technical characteristics.
One of its main features is that it is based on a pneumatic
technology, unlike most solutions on the market. This avoids
safety issues related to the introduction of electrical devices
in strong magnetic fields. In addition, being pneumatic-based,
Respiration Belt MR Transducer
Respiration Belt MR Sensor
Respiration Belt MR Transducer
Elastic belt and pouch
Auxiliary Connector Cable
Figure 1
www.brainproducts.com
Figure 2
page 1 of 2
Brain Products Press Release
the Respiration Belt MR is not a source of artifacts for the MR
imaging, thus preserving the highest data quality and ensuring
that no noise is induced on the MR recorded signal. Extensive
tests have been carried out with scanners from various
manufacturers with very satisfactory results.
Moreover, we developed our Respiration Belt MR with the aim
of having a device with great sensitivity which is able to
adequately follow different types of respiratory acts in a robust
way. Figure 2 shows a slow and deep respiration (black line)
and a faster and shallow respiration (red line) as measured
by the Respiration Belt MR: The Respiration Belt MR is able to
follow the dynamic of respiratory act over quite a wide range,
April 2013, Volume 46
showing a good sensitivity of the system. This makes of it a
powerful and sophisticated tool to obtain high quality respiratory
signals, and thus regressors for artifact correction, and also
to investigate interrelation between physiology and brain
organization more accurately. The higher sensitivity of the belt
to respiratory dynamics makes it easier and more effective to
compute higher order regressors describing fluctuations of
respiration over time.
We are convinced that the new Respiration Belt MR represents a
very useful instrument for advanced research over a wide range
of applications and we will be pleased to welcome any of your
further enquires.
Moriah E. Thomason, Brittany E. Burrows, John D.E. Gabrieli, Gary H.
Glover, Breath holding reveals differences in fMRI BOLD signal in children
and adults, NeuroImage, Volume 25, Issue 3, 15 April 2005, Pages 824837, ISSN 1053-8119, 10.1016/j.neuroimage.2004.12.026.
Birn, R. M., Murphy, K. and Bandettini, P. A. (2008), The effect of
respiration variations on independent component analysis results of
resting state functional connectivity. Hum. Brain Mapp., 29: 740–750. doi:
10.1002/hbm.20577.
Christopher G. Thomas, Richard A. Harshman, Ravi S. Menon, Noise
Reduction in BOLD-Based fMRI Using Component Analysis, NeuroImage,
Volume 17, Issue 3, November 2002, Pages 1521-1537, ISSN 1053-8119,
10.1006/nimg.2002.1200.
Glover, G. H., Li, T.-Q. and Ress, D. (2000), Image-based
method for retrospective correction of physiological motion
effects in fMRI: RETROICOR. Magn Reson Med, 44: 162–167. doi:
10.1002/1522-2594(200007)44:1<162::AID-MRM23>3.0.CO;2-E.
Erik B. Beall, Mark J. Lowe, The non-separability of physiologic noise in
functional connectivity MRI with spatial ICA at 3T, Journal of Neuroscience
Methods, Volume 191, Issue 2, 30 August 2010, Pages 263-276, ISSN 01650270, 10.1016/j.jneumeth.2010.06.024.
Cook R., Peel E., Shaw R.L., Senior C., 2007. The neuroimaging research
process from the participants’ perspective. International Journal of
Psychophysiology 63, 152–158.
www.brainproducts.com
page 2 of 2