| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Meakins Christie Laboratories, McGill University, and Notre Dame Hospital, University of Montreal, Montreal, Quebec, Canada H2L 4M1; and 2 Spinal Injuries Unit, Sahlgrenska Hospital, University of Göteborg, Göteborg, Sweden
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
APPENDIX
ACKNOWLEDGEMENTS
FOOTNOTES
REFERENCES
ABSTRACT 
Beck, Jennifer, Christer Sinderby, Lars Lindström, and
Alex Grassino. Diaphragm interference pattern EMG and compound muscle action potentials: effects of chest wall configuration. J. Appl. Physiol. 82(2): 520-530, 1997.
The effect of chest wall configuration on the diaphragm
electromyogram (EMGdi) was evaluated in five healthy subjects with an
esophageal electrode for both interference pattern EMGdi (voluntary
contractions) and electrically evoked diaphragm compound muscle action
potentials (CMAPs). Diaphragm CMAPs (both unilateral and bilateral)
were evaluated for the baseline-to-peak amplitude (Ampl), the time from
the onset of the CMAP to first peak (T1), root mean square (RMS), and
center frequency (CF) values of the CMAP power spectrum. CF values from
the interference pattern EMGdi power spectrum were also calculated. For
CMAPs obtained at an electrode position least influenced by variations
induced by electrode positioning, Ampl increased with diaphragm
shortening from functional residual capacity (FRC) to total lung
capacity (TLC) by 101 and 98% (unilateral and bilateral,
respectively). Bilateral CMAP RMS values increased 116% from FRC to
TLC. CMAP T1 values decreased with diaphragm shortening from FRC to TLC by 1.1 and 2.1 ms for the unilateral and bilateral stimulations, respectively, and CF increased for the bilateral diaphragm CMAPs with
diaphragm shortening. CF values from the interference pattern EMGdi did
not show any consistent change with chest wall configuration. Thus CF
values of the interference pattern EMGdi obtained with an esophageal
electrode can be considered reliable for physiological interpretation,
at any diaphragm length (if electrode positioning and signal
contamination are controlled for), contrary to the diaphragm CMAPs,
which are sensitive to changes in chest wall configuration. It is
speculated that the different results (over the effects of chest wall
configuration on interference pattern EMGdi and diaphragm CMAPs) may be
because of summation properties of the signals and how these influence
the EMG power spectrum.
spectral analysis; esophageal electrode; cross-correlation; crural
diaphragm; phrenic nerve stimulation
INTRODUCTION INVESTIGATIONS CONCERNING the effect of chest wall
configuration on esophageal recordings of the human diaphragm
electromyogram (EMGdi) have yielded conflicting results. For voluntary
contractions of the diaphragm, it has been shown that the effect of
chest wall configuration on the EMGdi center frequency (CF) is
negligible, if muscle-to-electrode distance filtering effects and
signal contamination are controlled for (4). With respect to
electrically evoked diaphragm compound muscle action potentials
(CMAPs), Gandevia and McKenzie (10) observed a systematic increase in
the amplitude (Ampl) and area of the CMAPs with increasing lung volume
after supramaximal phrenic nerve stimulation. Their data also suggested that there was an increase in the frequency content of the CMAP, by a
reduction in the time to reach the peak of the CMAP with diaphragm
shortening. Others have also observed that esophageal recordings of
diaphragm CMAPs are influenced by lung volume (7, 16, 22).
The voluntary EMGdi and the diaphragm CMAPs are myoelectric signals
that differ with respect to the summation of the motor unit action
potentials contributing to the signal. The voluntary signal, hereafter
referred to as the interference pattern EMGdi, represents the summated
electrical activity generated by asynchronously firing crural diaphragm
motor units; the diaphragm CMAP represents the summated synchronized
electrical activity generated by all motor units after a supramaximal
stimulus of the phrenic nerve. Perhaps the conflicting reports over the
effects of lung volume on the diaphragm CMAPs and the interference
pattern EMGdi are due to the differences in the nature of the two
signals. As well, one cannot exclude the possibility that differences
in methodology may be responsible for the inconsistent results. To our
knowledge, no studies have been published that directly compare the
effects of chest wall configuration on voluntarily and electrically
evoked EMGdi, as obtained in the same subjects, with the same
methodology.
The purpose of the present study was to evaluate the effects of chest
wall configuration on esophageal recordings of EMGdi 1) after supramaximal, unilateral,
and bilateral phrenic nerve stimulation (diaphragm CMAPs); and
2) during voluntary contractions of
the diaphragm (interference pattern EMGdi), with the same methodology. We expected that the results would help to resolve the issue
surrounding the effects of lung volume on esophageal recordings of
EMGdi. The question is relevant for determining the validity of
esophageal electrodes in the measurement of EMGdi (CF or the diaphragm
CMAPs), for example, in the evaluation of the acute development of
diaphragmatic fatigue, independent of diaphragm length.
METHODS Subjects
Signal Acquisition
Fig. 1.
Experimental setup. Left: esophageal
electrode used in present study. Most caudal pair of electrodes was
arbitrarily designated electrode pair
1, and most cephalad, electrode pair
4. Thin-pressure lumen was included in catheter that
allowed measurement of gastric pressure (Pga).
Center: Konno and Mead diagram
[Ref. 13; y-axis, rib cage (RC)
displacement; x-axis, abdominal (Ab)
displacement] showing 4 chest wall configurations examined in
present study: functional residual capacity (FRC) on relaxation curve,
FRC with belly in (FRCbin), 50% of inspiratory capacity
(IC50), and total lung capacity
(TLC). EMG, electromyogram, Pes, esophageal pressure; Pdi,
transdiaphragmatic pressure;
Pdimax, maximal voluntary Pdi.
[View Larger Version of this Image (20K GIF file)]
Bipolar surface recordings of the diaphragm CMAPs were obtained for both the left and right sides to monitor supramaximality of the phrenic nerve stimulation. The surface diaphragm CMAPs were recorded via two pairs of surface silver-silver chloride ECG electrodes (diameter = 1 cm; FC24, Graphic Controls) placed over the seventh or eighth intercostal space, anterior to the midaxillary line, near the costal margin. The interelectrode distance was 3-5 cm.
A Teflon tube (diameter = 1 mm) was placed inside the silicone catheter and a 5-cm-long 1.5-cm-diameter latex balloon was attached 5 cm distal to the most distal ring to allow the measurement of gastric pressure (Pga) at the distal end of the esophageal electrode. Esophageal pressure (Pes) was measured by a separate catheter and standard balloon system (18). The pressure catheters were connected to two differential pressure transducers (±150 cmH2O; Validyne, Northridge, CA), allowing the measurement of Pes and Pga changes relative to atmospheric pressure. Transdiaphragmatic pressure (Pdi) was calculated by subtracting Pes from Pga and was displayed to the subject on a storage oscilloscope (Tektronix 5103N) (Fig. 1, right).
EMGdi signals from each of the four pairs of electrodes on the esophageal catheter were amplified and band-pass filtered between 32 and 800 Hz (electromyograph model TE4, Teca, Pleasantville, NY). EMGdi signals from the esophageal electrode were displayed and monitored on a storage oscilloscope (model 1604, Gould). The ECG signals were amplified and band-pass filtered between 16 and 32 Hz (model 8811A bioelectric amplifier, Hewlett-Packard). The EMGdi signals from the esophageal electrode and the ECG signal were acquired and digitized by an analog-to-digital converter (Data translation 2801) with 12-bit resolution, at a sampling frequency of 2 kHz. The EMGdi signals obtained from the chest surface electrodes were amplified and band-pass filtered between 16 and 1,600 Hz (electromyograph model TE4, Teca) and were displayed to the investigators on a computer monitor.
Evaluation of Chest Wall Configuration
Chest wall configuration was assessed throughout the experiment by the method of Konno and Mead (13). Two respiratory-inductive plethysmography bands were used (Respitrace, Ambulatory Monitoring,) to evaluate rib cage (RC) displacemnent and abdominal (Ab) displacement. The RC band was placed around the upper portion of the thorax, vertically centered over the nipples, whereas the upper edge of the Ab band was placed around the abdomen at the level of the umbilicus. The RC signals were amplified and displayed on the vertical axis and the Ab signals on the horizontal axis of a storage oscilloscope (Tektronix, 5103N) (Fig. 1, center).Experimental Protocol
Subjects were studied while they were seated upright in a chair. Respitrace bands were positioned on the subjects and secured in place by a surgical bandage placed over the thorax. Subjects were trained to perform isovolume maneuvers and relaxation curves with visual feedback from the oscilloscope. Once familiar with these maneuvers, subjects practiced reaching defined points on the obtained Konno and Mead diagram (see Ref. 13; Fig. 1, center).After the training session, the esophageal electrode was passed through the nose, swallowed, and positioned at the level of the esophageal hiatus such that the four electrode pairs received the highest amplitude at total lung capacity (TLC). The electrode was then fixed externally at the nose. The Pes catheter was then also passed through the nose, and its position was confirmed by the occlusion test (2) and then fixed at the nose. After positioning of the catheters, a maximal voluntary Pdi (Pdimax) maneuver was performed (combined Mueller-expulsive) at functional residual capacity (FRC). The highest of three reproducible values was recorded, and 20% of the Pdimax at FRC was displayed on the oscilloscope as a target for the subject. With the catheters in place, subjects were asked to perform a relaxation maneuver from TLC to FRC and a series of isovolume maneuvers at FRC, 50% of inspiratory capacity, and at TLC (Fig. 1, center). To ensure that posture and the position of the Respitrace bands remained constant, these maneuvers were repeated throughout the experiment.
For both the stimulation protocol and the voluntary contractions protocol, four chest wall configurations were evaluated: FRC with the belly in (FRCbin), FRC on the relaxation curve, 50% of inspiratory capacity (IC50), and TLC (see Fig. 1, center). We assumed that the diaphragm was longer at FRCbin than at FRC (12) and that further shortening occurs from FRC to TLC (11). The phrenic nerve stimulation protocol was performed before the voluntary contractions, with a 20-min rest period between the two protocols.
Voluntary contractions. The protocol for the voluntary contractions was the following: subjects were asked to reach a predetermined point on the Konno and Mead diagram (Ref. 13; marked on the oscilloscope) and, while keeping the beam of the scope at the target point (i.e., maintaining the same chest wall configuration), performed a voluntary contraction of the diaphragm, maintaining 20% of the Pdimax obtained at FRC. Each contraction lasted 8-10 s and was repeated randomly five times for each chest wall configuration. A rest period was allowed between contractions (1-5 min). Phrenic nerve stimulation. For the phrenic nerve stimulation protocol, subjects were asked to reach a predetermined point on the Konno and Mead diagram (13). Once the subject relaxed at the particular chest wall configuration (against a closed glottis), the phrenic nerves were stimulated transcutaneously and either unilaterally (left) or bilaterally at the neck (19) with 0.1-ms square-wave pulses delivered by a dual-output constant-current stimulator (model N S6, Teca). Current intensities in the range of 1-50 mA were delivered to produce maximal electrical responses from the diaphragm. Maximal responses were maintained by increasing the current intensity by 40% over that necessary to produce a maximal response. At each chest wall configuration, a series of 10 twitches was recorded. In one subject, a series of 12 supramaximal phrenic nerve stimulations was performed during a passive expiration from TLC to FRC to repeat the experimental protocol of Gandevia and McKenzie (Ref. 10; cf. Fig. 2).
Signal Analysis
Signal analysis was performed off-line for both the interference pattern EMGdi and the diaphragm CMAPs. Common to the analysis of both types of signals was the control of esophageal electrode positioning with respect to the diaphragm. Locating the position of the diaphragm along the multiple electrode array is essential for reliable interpretation of esophageal recordings of EMGdi (3). With a multiple array of electrodes (interelectrode distance of 10 mm), control of electrode positioning can be achieved by locating the electrode pair that lies closest to the center of the electrically active region of the diaphragm (EARdi) and using it as a reference (3). Once the electrode pair closest to the center of the EARdi is determined, the position of the electrode pairs cephalad and caudal to this point can be described in terms of distance away from the EARdi center. In the present study, for each interference pattern EMGdi segment and for each diaphragm CMAP, control of electrode positioning was considered in the analysis (see below). Interference pattern EMGdi analysis. Interference pattern EMGdi signals were automatically processed with computer algorithms previously described in detail (21). The algorithms allow the selection of segments between successive QRS complexes of the ECGs and exclude interference pattern EMGdi signals, which are contaminated by noise, motion artifacts, esophageal peristalsis, and ECG P or T waves. Interference pattern EMGdi segments from all four electrode pairs were automatically selected between the ECG QRS complexes as a percentage of the R-R interval (50-75%). In the present study, the relative position of the EARdi center with respect to the electrode pairs was determined for each EMGdi segment. To locate the electrode pair closest to the center of the EARdi, computer algorithms performing cross-correlation analysis (3) were used (Fig. 2). Cross-correlograms were calculated for signals from electrode pairs with an interpair distance of 20 mm (center to center), i.e., signals from electrode pair 1 vs. 3 and electrode pair 2 vs. 4 (Ref. 4, see Fig. 2, right). For a bipolar electrode arranged perpendicularly to the muscle fiber direction, as in the case of esophageal recordings of EMGdi, the cross-correlogram for signals obtained on either side of the diaphragm should peak at negative values at the 0-ms time offset (3). The electrode pair centered between the two most negatively correlated pairs is the electrode pair closest to the EARdi center. In the example presented in Fig. 2 (right), electrode pair 2 vs. 4 were the most negatively correlated (r =
0.69), and therefore
electrode pair 3 is the electrode pair
closest to the center of the
EARdi. Once the electrode pair
closest to the center of the EARdi
is determined (electrode pair 3),
the position of the electrode pairs cephalad and caudal to this point
can be described in terms of "distance away" from the electrode
pair closest to the EARdi center.
In this example (Fig. 2), electrode pair
4 is 10 mm cephalad to the EARdi center, and
electrode pairs 1 and
2 are 20 and 10 mm caudal to the
electrode pair closest to the
EARdi center, respectively. All
interference pattern EMG segments from the four electrode pairs
selected between the ECGs were arranged with respect to distance away
from the electrode pair closest to the
EARdi center.
After determination of the EARdi
center with respect to the electrode pairs, the power spectrum CF was
calculated, and signal contamination was evaluated for each EMGdi
segment (21). From the selected EMGdi segments (between the ECGs), the
direct current levels and slopes were determined by linear-regression
analysis and removed. The first and last zero crossings of the
segment were then determined, and the tails of the EMGdi
segment were zero padded to fit the segments for a fast Fourier
transform (FFT) of 1,024 points. The time domain EMGdi segment was
converted into the frequency domain by FFT (6), and the power spectra
were calculated. The power spectra obtained were then evaluated by four
signal contamination indexes: the signal-to-motion artifact (SM) ratio,
the signal-to-noise (SN) ratio, the drop in power density of the
spectrum (DP) ratio, and a spectral deformation (
) ratio (21). SM,
SN, and DP are all expressed in decibels (dB), whereas is expressed
in relative units. Only signals fulfilling the recommended levels of SM > 12 dB, SN > 15 dB, DP > 30 dB, and < 1.4 were included
in the analysis (21). The CF was calculated from the EMGdi power
spectrum.
In each subject, a mean (±SD) CF was calculated for each of the
different voluntary diaphragmatic contractions, for each chest wall
configuration, and for each electrode pair position. To ensure that the
diaphragmatic contractions were nonfatiguing, CF values were evaluated
with respect to time for each contraction. To evaluate the effects of
chest wall configuration on CF values, a mean (±SD) CF for the five
subjects was calculated for each electrode position and for each chest
wall configuration. As well, the change in CF values from FRC (and the
%change from FRC) were calculated for each of the five subjects at the
different electrode pair positions.
The root mean square (RMS) of the interference pattern EMGdi power
spectrum was not analyzed or evaluated in the present study because of
changes in diaphragm activation with changes in chest wall
configuration.
Diaphragm CMAPs.
The electrode pair closest to the diaphragm was determined by visually
inspecting the diaphragm CMAPs. In Fig. 3,
a series of seven similar diaphragm CMAPs is presented (superimposed). In the example presented (Fig. 3), reversal of signal polarity occurred
between electrode pairs 2 and
4, indicating that the EARdi center was between them and,
hence, closest to electrode pair 3. As
for the interference pattern EMGdi, electrode pair 4 in this case can be labeled as 10 mm cephalad
to the EARdi center, electrode pair 2 as 10 mm caudal to
the center of the EARdi, and electrode pair 1 as 20 mm caudal to
the center of the EARdi. The position of the EARdi center was
also confirmed by analyzing the interference pattern EMGdi (with
cross-correlation) from the brief voluntary contraction that was
present at the beginning of the diaphragm CMAP files at the higher lung
volumes.
At each chest wall configuration, CMAPs from the electrode pair that was the least influenced by variations caused by electrode positioning (usually located 10-20 mm caudal to the electrode pair closest to the EARdi center) were selected for analysis. Hereafter, this selected electrode pair is referred to as the "least influenced electrode pair." Diaphragm CMAPs from the least influenced electrode pair were evaluated for baseline-to-peak amplitude and the time from CMAP onset to the first peak (T1) (see Fig. 3), unless the signals were influenced by the ECG. The change in Ampl and T1 values from FRC (and the %change from FRC) were calculated for each of the four subjects. It was anticipated that frequency domain analysis of the diaphragm CMAPs would also be performed; however, this was possible in only two of the five subjects for the unilateral signals and in three out of the five subjects for the bilateral signals. The frequency domain analysis of the diaphragm CMAPs requires that the signal returns to baseline at the end of the action potential, and this was not fulfilled in the excluded subjects. We noted that the swing in Pes began 18-20 ms after the onset of the diaphragm CMAP, which probably caused motion artifacts in the signal and was the source of the signal not returning to baseline. In the subjects whose diaphragm CMAPs returned to baseline, frequency domain analysis was performed for each diaphragm CMAP in the run (uncontaminated by the ECG). A 512-ms segment was selected, starting from the beginning of the diaphragm CMAP. The direct current value obtained at the starting point of the signal was subtracted from the entire segment. When the diaphragm CMAP returned to baseline (i.e., reached a second zero crossing), zeroes were added to the signal to complete the 512-ms array for the FFT. The time domain EMGdi segment was then converted into the frequency domain by FFT (6), and the power spectrum was calculated. CF was calculated from the power spectrum. The RMS was calculated as the square root of the area under the power spectrum. In contrast to the interference pattern EMGdi, the RMS could be evaluated for the CMAPs because supramaximal stimulation ensures constant neural drive to the diaphragm. An average CF and RMS value (mean ± SD) was calculated for the group for the least influenced electrode pair at each chest wall configuration for both the bilateral and unilateral stimulations (in those subjects whose diaphragm CMAPs returned to baseline; n = 3). As well, the change in CF (Hz) and RMS [arbitrary units (au)] values with respect to the values at FRC (and the %change from FRC) were calculated for each of the three subjects.
Statistical Analysis
The effect of changing diaphragm length on interference pattern EMGdi CF values and diaphragm CMAPs T1, Ampl, CF, and RMS values was estimated by simple linear-regression analysis from chest wall configurations FRCbin to TLC because diaphragm shortening was expected to occur from FRCbin to TLC (11, 24). If linear-regression analysis did not reach statistical significance, two-way analysis of variance (ANOVA) and post hoc multiple comparisons were used to evaluate the variabilty of the parameters with changes in chest wall configuration. All statistical tests were performed with Sigmastat (Jandel Scientific). P < 0.05 was chosen to be the level of statistical significance.RESULTS
Effects of Chest Wall Configuration on Interference Pattern EMGdi
CF values for each subject and for each contraction were evaluated with respect to time, and there was no evidence for reductions in CF during the contractions, indicating that the contractions were nonfatiguing. Figure 4 demonstrates the mean (±SD) CF value in hertz (y-axis) for each chest wall configuration at each electrode position for one subject. At a given chest wall configuration, the variability in CF values with changes in electrode positioning was high, the maximum range being ~50 Hz along the entire electrode array. As previously described (3), CF values were the highest at the EARdi center and progressively decreased with increasing muscle-to-electrode distance. However, for a given electrode position, CF values showed no consistent change with changes in chest wall configuration.
For the group of five subjects (Fig. 5),
EMGdi mean (±SD) CF values (y-axis) showed
no consistent changes with chest wall configuration
(x-axis) for any electrode position
(top to
bottom). The maximum range in mean
CF values that was observed (at any electrode position) from one chest
wall configuration to the next was 7 Hz, ~5% of the CF values
obtained at FRC. The mean (±SD) change in CF values from FRC (in Hz
and %) are presented in Table 1 for all
electrode pair positions.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Effects of Chest Wall Configuration on Diaphragm CMAPs
For the diaphragm CMAPs, Fig. 6 demonstrates the raw signals for the four chest wall configurations (top to bottom) and for each electrode position (left to right) obtained in one subject after supramaximal bilateral phrenic nerve stimulation. The scaling of the diaphragm CMAPs are the same, and signals are plotted in terms of arbitrary units. For any given chest wall configuration, the effects of electrode positioning on the amplitude, waveform, and polarity of the diaphragm CMAPs can be observed (left to right). It was observed that at the EARdi center, Ampl was strongly reduced. In moving away from the EARdi center, diaphragm CMAPs became larger, and then, in some subjects, again smaller due to distance-filtering effects. The extent of the cancellation effects at the EARdi center and the filtering effects due to increasing muscle-to-electrode distance varied from subject to subject. For a given electrode position, a trend for an increase in CMAP Ampl with increasing lung volume (bottom to top) can be seen.
Diaphragm CMAPs obtained in all but one of the subjects were used in the analysis. The CMAPs obtained in subject 1 indicated that at high lung volumes, the diaphragm moved off the electrode (consistently for the unilateral and bilateral stimulations), making determination of the EARdi center impossible. For this reason, the data obtained from subject 1 were not included in the group mean calculations; however, the general trend of an increase in Ampl with diaphragm shortening was observed in subject 1.
The Ampl parameters, Ampl (Fig.
7A) and
RMS (Fig. 7B), and the frequency-related parameters are
plotted in Fig. 7 for the group (mean ± SD) for the different chest
wall configurations (x-axis) for both
bilateral and unilateral supramaximal phrenic nerve stimulations. For
each subject, data were obtained from the least influenced electrode
pair (see METHODS). Ampl
consistently increased with diaphragm shortening from FRC to TLC by
101% (range 42-198%) and 98% (range 52-125%) for the
unilateral and bilateral stimulations, respectively. There was a slight
increase in Ampl from FRC to FRCbin for both types of
stimulations, although the increase was not significant.
Linear-regression analysis indicated that Ampl for both the unilateral
and bilateral stimulations increased with diaphragm shortening
(P < 0.05) from FRCbin
to TLC. RMS values (plotted only for bilateral stimulations in subjects
in whom frequency domain analysis was possible;
n = 3) increased from
FRCbin to TLC (P < 0.05)
by 116%, in accordance with the Ampl results. The mean (±SD)
change in Ampl and RMS values from FRC (in au and %) are presented in
Table 1.
Mean (±SD) T1 values for the group (Fig. 7C), a reflection of the frequency content of the signal, showed consistent reductions, with diaphragm shortening from FRC to TLC, decreasing on average by 1.1 and 2.1 ms for the unilateral and bilateral stimulations, respectively; there was a slight decrease in T1 with diaphragm lengthening from FRC to FRCbin. The trend for a decrease in T1 with diaphragm shortening was not significant; however, two-way ANOVA indicated that T1 varied significantly with chest wall configuration (±SD) change in T1 values from FRC (in ms and %) are presented in Table 1 for the bilateral and unilateral stimulations. Multiple comparisons indicated that the T1 values at FRC were statistically different from the T1 values at TLC for both the unilateral and bilateral stimulations and from FRCbin for the unilateral stimulations. CF values of the CMAPs (Fig. 7D) increased significantly from FRCbin to TLC by 21% (P < 0.05) (see Table 1).
Figure 8 represents, in one subject, the
diaphragm CMAPs obtained after bilateral stimulation during a passive
expiration (~10 s long) from TLC to FRC. This part of the study was
performed to demonstrate the relationship between lung volume and the
diaphragm CMAP in the same manner as that performed by Gandevia and
McKenzie (10). Only the signals as they come out for the different
electrode pairs are displayed, and the center of the
EARdi was not accounted for. On
the basis of observation of the signals from all electrode pairs, the
finding of an increase in Ampl with increased lung volume is confirmed,
as well as a reduction in T1. The least filtered electrode
pair in this case is electrode pair 4,
in which an increase in Ampl and a reduction in T1 are observed. Figure
8 emphasizes the sensitivity of the shape, polarity, and Ampl of
diaphragm CMAPs to electrode positioning.
DISCUSSION
The present study demonstrates that diaphragm CMAPs measured with an esophageal electrode increased in baseline-to-peak amplitude and RMS and are reduced in T1 and increased in CF with increasing lung volume from FRC to TLC. The interference pattern EMGdi CF, however, measured with the same esophageal electrode and in the same subjects, shows no consistent changes related to chest wall configuration, when signal contamination and electrode positioning are controlled for.
The results we obtained concerning the volume-related effects on the amplitude of esophageal recordings of diaphragm CMAPs are consistent with those reported previously by others (7, 10, 16, 22). None of the above-mentioned studies reported the effects of chest wall configuration on the frequency content of the diaphragm CMAPs. Mead (16) was the first to demonstrate a systematic increase in the diaphragm CMAPs both when lung volume was increased and when, at a constant lung volume, the abdomen was expanded and the RC was simultaneously compressed. He commented at the end of his discussion that he and his colleagues had inconstant stimulation that was submaximal and that no evidence as to the reliability of the esophageal electrodes could be deduced from the experiments. Gandevia and McKenzie (10) later confirmed the findings of an increase in CMAP amplitude with increasing lung volume (when stimulation remained supramaximal), as previously reported by Mead (16). These authors reported that the amplitude and area of the diaphragm CMAPs consistently increased with increasing lung volume (amplitude and area increased 2.5 and 5 times from FRC to TLC, respectively). Smith and Bellemare (22) reported as much as a fourfold increase in the amplitude of the diaphragm potentials as recorded with an esophageal electrode from residual volume to TLC. On the basis of their own findings, Gandevia and McKenzie (10) stated that esophageal recordings of diaphragmatic EMG are unreliable because, under conditions of constant "neural drive" (i.e., supramaximal phrenic nerve stimulation), diaphragm CMAPs vary with chest wall configuration. In an attempt to minimize the effect of lung volume on the diaphragm CMAP, Daubenspeck et al. (7) rectified and integrated the EMG signals from each of seven pairs of electrodes of a multiple-electrode array and summed the values from each of the pairs to give an approximation of the total activity over the span of the array. With this approach, they found that the diaphragm CMAP amplitude still increased by up to 1.6 times from FRC to TLC. The observation of an increase in diaphragm CMAP with increasing lung volume is not altered when different variations of the esophageal electrode were used, i.e., when one or several electrode pairs were used or when the esophageal electrode was anchored with a balloon at the cardia or externally at the nose.
A study performed by Delhez (8) demonstrated that latencies (time from stimulus to onset of CMAP) were different for right- and left-side unilateral stimulations. He also showed (9) that the position of the reversal in CMAP polarity (in terms of distance from the nares) was different for the right and left unilateral stimulations. McKenzie and Gandevia (15) described that the optimal site for recording a diaphragm CMAP with an esophageal electrode was ~1-2 cm higher for left-side unilateral stimulation than for right-side unilateral stimulation and proposed this to be due to different positions of the left and right crural diaphragm fibers with respect to the electrode. These findings suggest that bilateral stimulation would result in a CMAP that represents the summation of the left and right sides and that phase cancellation of the potentials would occur due to different arrival times of the potentials at the electrode pairs. In the present study, however, no attempts were made to evaluate the effects of left- and right-side contributions to the esophageal CMAPs for the bilateral stimulations. It was also not possible to compare optimal recording positions for the diaphragm CMAPs elicited for left and right unilateral stimulations because this would require that the electrode-array position with respect to the left and right sides be exactly the same; as we have demonstrated (Figs. 6 and 8), the diaphragm CMAP amplitude, shape, and polarity are very sensitive to electrode positioning.
Gandevia and McKenzie (10) and Daubenspeck et al. (7) suggested that the increase in diaphragm CMAP amplitude with increasing lung volume was due to reductions in the radial distance between the crural diaphragm and the electrodes. A reduction in muscle-to-electrode distance results in an increase in the amplitude and the frequency content of the signal (14). Muscle-to-electrode distance-filtering effects have been evaluated for esophageal recordings of the voluntary EMGdi in terms of axial displacement of the diaphragm along a multiple-electrode array (4) but have not been described for changes in radial distance. In the present study, the observed increases in CMAP Ampl and CF with increasing lung volume agree with the concept of reductions in radial distance during these maneuvers. However, the results of the present study for the voluntary EMGdi did not demonstrate an increase in CF values (for any electrode pair position) when lung volume was increased and are in agreement with a previous study (4).
It cannot be excluded that the radial distance of a contracting diaphragm (during static voluntary contractions at a constant force) and of a relaxed diaphragm during the evoked CMAPs may behave differently with respect to lung volume. When the diaphragm contracts at a given force, it may have a more or less constant radial distance around the electrode, independent of lung volume, whereas the radial distance of a relaxed diaphragm may decrease with increasing lung volume. In addition, perhaps with the belly-in maneuver (FRCbin), the radial distance around the electrode is not increased from FRC to FRCbin and maybe is even reduced. Therefore, it could be that the concept of changes in radial distance is one explanation for the changes seen in diaphragm CMAP Ampl with increasing lung volume.
With respect to voluntary contractions of the human diaphragm, few reports are available that have specifically aimed to investigate the effect of lung volume and chest wall configuration on the CF of EMGdi. Schweitzer et al. (20) reported an observation that CF values increased during the early part of inspiration and attributed their findings to recruitment of fibers with faster action potential conduction velocities. Aldrich et al. (1) reported an increase in CF values with inspiratory duration; however, the observation was inconsistent between patients and healthy subjects. The influence of signal contamination on the EMGdi was considered in both of these studies; however, their methods of detecting artifacts in the signal were not entirely objective. In addition, Aldrich et al. (1) did not control for changes in axial displacement of the diaphragm with respect to their electrode, a factor known to strongly filter the interference pattern EMGdi CF values (4). We have previously shown that an electrode fixed externally at the nose can move up to 4 cm with changes in lung volume from FRC to TLC (4). Both Aldrich et al. (1) and Schweitzer et al. (20) studied the changes in CF values during the early phase of a dynamic unloaded inspiration. Weinberg et al. (23), who studied the interference pattern EMGdi in spinal cord-injured patients with an anchored esophageal electrode, reported that the majority of EMG power spectra are deformed early on during unloaded inspirations (as quantified by the power spectrum deformation index) and that the relative contribution of the contaminating signals in the interference pattern EMG decreases proportionally throughout the first two-thirds of the inspiratory capacity. We believe that previous observations of increasing CF values during early inspiration (1, 20) may have been influenced by changes in power spectrum quality and changes in the position of the diaphragm with respect to the recording electrodes during the early part of inspiration. When both signal contamination and electrode positioning effects were controlled and static contractions of the diaphragm were performed at different chest wall configurations, no consistent changes in CF were found in the present or in a previous study (4).
The findings of the present study in five subjects, and those of a previous one in four additional subjects (4), indicate that the CF values of the EMGdi power spectrum are not affected by changes in chest wall configuration in healthy subjects. It should also be pointed out that the observations of an increase in CMAP Ampl with increasing lung volume seem to be valid findings because they have been demonstrated by others (7, 10, 16, 22) and confirmed by us in the present study, even with differences in methodology. The findings obtained in the healthy subjects studied in the present work (i.e., no change in CF for interference pattern EMG but increase in CMAP Ampl with lung volume) should be reevaluated for the interference pattern EMGdi in patients with neuromuscular pathology, in whom the motor units are enlarged and the interference pattern displays more distinct motor unit action potentials, similar to the CMAPs.
The conflicting results over the effects of chest wall configuration on esophageal recordings of interference pattern EMGdi and diaphragm CMAPs demonstrated in the present study may be due to differences in the nature of the two signals. The CMAP is a deterministic signal that is evoked by a synchronized stimulus of all motor units, whereas the spontaneous EMG is a stochastic signal generated by the activation of any number of motor units (17). Due to the differences in the activation patterns, the power spectrum characteristics are different as well. According to Lindström and Magnusson (14), the influence of summation effects on the power spectrum can be described by
|
is the angular frequency (or 2
× frequency), and
is a measure of the time
dispersion of the individual potentials contributing to the signal (SD)
and is dependent on the spread of innervation points. The summation
function indicates that the power spectrum is influenced by
N, the number of fibers recruited and
contributing to the signal, on a square-law basis in the low- frequency
range of the spectrum, in which the amount of power is proportional to
the number of fibers squared. In the high-frequency range of the
spectrum, the amount of power is proportional on a linear basis to
N. The summation function is also
highly dependent on and the
spread in innervation points
(x).
When all motor units are activated synchronously during a supramaximal
stimulation (CMAP), the muscle can be considered as one large motor
unit (N is large and constant), with a
given x. With changes in
muscle length, one can expect a reduction in the size of
x, but no changes in
N, during supramaximal stimulation.
This will result in the power spectrum shifting to higher frequencies,
accounting for the increase in frequency content for the stimulated
signal with diaphragm shortening (or the reduction in T1 with diaphragm
shortening). With respect to the voluntary activity, where the muscle
does not behave as a single motor unit (and a number of motor units may
be active at a given time), changes in muscle length may change
recruitment patterns, and hence may vary
N, resulting in a complicated
interplay (in terms of frequency content of the signal) between the two
paramaters (x and
N) of the summation function
described above.
The effects of changing chest wall configuration on the Ampl of the
CMAP can be suggested to be due to changes in
x with changes in muscle
length and is described mathematically in the
Appendix. Generally speaking, the Ampl
of a CMAP is proportional to N, is
inversely proportional to ,
and hence muscle length, and is almost insensitive to the action
potential conduction velocity (see
Appendix).
The theory that changes in muscle length were the cause of the increased Ampl and frequency content of the diaphragm CMAPs is supported by the Ampl, RMS, T1, and CF values that we obtained from FRC to TLC; however, FRCbin, in which the diaphragm is expected to be longer than at FRC, did not follow the mathematically predicted behavior of the signal. We can only speculate that this may be because 1) the crural diaphragm at FRCbin is, in fact, not longer than at FRC; or 2) our observations are not due to changes in diaphragm length but more likely due to changes in the radial distance of the relaxed diaphragm (at different lung volumes), as mentioned above for the CMAPs. We could hypothesize that during a belly-in maneuver the increased abdominal pressure results in a reduced muscle-to-electrode (radial) distance and hence an increase in the CMAP Ampl and RMS, a reduction in T1, and an increase in CF.
In conclusion, the present study demonstrated that bipolar esophageal recordings of diaphragm CMAPs consistently increase in amplitude and frequency content with increasing lung volume. On the other hand, CF values of the interference pattern EMGdi obtained with a bipolar esophageal electrode show no consistent changes with changes in chest wall configuration. The results of the present study, therefore, resolve previous concerns about the reliablility of esophageal recordings of the interference pattern EMGdi in detecting early signs of diaphragmatic fatigue (5). This suggests that power spectrum analysis of the crural diaphragm interference pattern EMG has the potential to be used as an early indicator of acute diaphragm fatigue in the clinical environment, independent of lung volume. Although we have resolved the issue concerning the influence of chest wall configuration on the interference pattern EMGdi CF values, a mathematical model that describes why the two signals behave differently with respect to changes in chest wall configuration remains to be developed.
ACKNOWLEDGEMENTS
The authors thank Drs. F. Bellemare and M. Petitjean for expertise and assistance with this study.
FOOTNOTES
Address for reprint requests: J. Beck, Pulmonary Function Laboratory (I-2158), Hôpital Notre-Dame, 1560 Sherbrooke St. East, Montreal, Quebec, Canada, H2L 4M1.
Received 16 April 1996; accepted in final form 18 September 1996.
APPENDIX
CMAP Ampl Increases With Muscle Shortening
The power spectrum [
W()]
of diaphragmatic myoelectric signals lead off with a differential
esophageal electrode is well described (4) by the following
formula
![[&Dgr;<IT>W</IT> (&ohgr;)] = <IT>v</IT><SUP>2</SUP>(&ohgr;/<IT>v</IT>)[<IT>N</IT> + <IT>N</IT> (<IT>N</IT> − 1) exp(−&ohgr;<SUP>2</SUP>&sfgr;<SUP>2</SUP><SUB>&tgr;</SUB>)]](/content/vol82/issue2/fulltext/520/img002.gif)
|
![[<IT>K</IT><SUB>0</SUB>(&ohgr;<IT>h</IT><SUB>1</SUB>/<IT>v</IT>) − <IT>K</IT><SUB>0</SUB>(&ohgr;<IT>h</IT><SUB>2</SUB> /<IT>v</IT>)]<SUP>2</SUP>/<IT>K</IT><SUP> 2</SUP><SUB>0</SUB>(&ohgr;<IT>a</IT>/<IT>v</IT>)](/content/vol82/issue2/fulltext/520/img003.gif)
|
(A1) |
is the angular frequency (2 times the frequency in Hz),
v is the propagation velocity of the
muscle fiber action potentials, N is
the number of signal sources contributing to the total myoelectric
signal, is the time
dispersion of the individual source signals,
Ko( ) is the modified Bessel
function of the second kind and order zero describing the distance
filtering of signals from line sources,
h1 and
h2 are the
distances from the electrode plates to the sources under consideration,
and a is the muscle fiber radius.
Supramaximal phrenic nerve stimulation results in a myoelectric
compound signal from virtually all muscle fibers in the diaphragm; i.e., N is very large. Also,
is very large because it
consists of both the dispersion within motor units and dispersion
between units. Assuming that
is caused mainly by the
spread in innervation points, we can write
|
(A2) |
unit is the SD of the
distribution of innervation points within a motor unit, and
muscle is the SD of the
distribution of centers of motor units. The two spreading processes are
assumed to be uncorrelated. Because of the large time dispersion of
contributions to the compound signal, the factor exponent
22
will dominate the high-frequency roll-off of the power spectrum rather
than the distance-filtering effect of the Bessel functions. The
difference between the Bessel functions still suppresses low-frequency
contributions and can be approximately described by a function that
increases weakly with frequency, e.g.
![[<IT>K</IT><SUB>0</SUB>(&ohgr;<IT>h</IT><SUB>1</SUB>/<IT>v</IT>) − <IT>K</IT><SUB>0</SUB>(&ohgr;<IT>h</IT><SUB>2</SUB>/<IT>v</IT>)]<SUP>2</SUP>/K<SUP>2</SUP><SUB>0</SUB>(&ohgr;<IT>a</IT>/<IT>v</IT>) ≈ &ohgr;<IT>h</IT><SUB>0</SUB>/<IT>v</IT>](/content/vol82/issue2/fulltext/520/img005.gif)
|
(A3) |
![[&Dgr;<IT>W</IT>(&ohgr;)] = <IT>v</IT><SUP>−2</SUP>(&ohgr;/<IT>v</IT>)<IT>N</IT> <SUP>2</SUP> exp(−&ohgr;<SUP>2</SUP>&sfgr;<SUP>2</SUP><SUB>&tgr;</SUB>)(&ohgr;<IT>h</IT><SUB>0</SUB>/<IT>v</IT>)](/content/vol82/issue2/fulltext/520/img006.gif)
|
(A4) |
|
(A5) |
|
(A6) |
Because the summation-filtering effects override those of the distance filtering, the duration is dominated by the influence of the time dispersion of the signals contributing to the compound signal. Thus
|
(A7) |
Putting all parts together, we find
|
 d&ohgr;](/content/vol82/issue2/fulltext/520/img011.gif)
|
(A8) |
|
(A9) |
|
(A10a) |
|
(A10b) |
unit and
muscle change with changing
muscle length (L) and hence
|
(A11) |
The amplitude of the compound signal is thus approximately proportional to N, approximately insensitive to v, and approximately inversely proportional to the L.
REFERENCES
| 1. |
Aldrich, T. K.,
J. M. Adams,
N. S. Aurora,
and
D. F. Rochester.
Power spectral analysis of the diaphragm electromyogram.
J. Appl. Physiol.
54:
1579-1584,
1983.
|
| 2. | Baydur, A., P. K. Behrakis, W. A. Zin, M. Jaeger, and J. Milic-Emili. A simple method for assessing the validity of the esophageal balloon technique. Am. Rev. Respir. Dis. 126: 788-791, 1982. [Medline] |
| 3. |
Beck, J.,
C. Sinderby,
L. Lindström,
and
A. Grassino.
Influence of bipolar electrode positioning on measurements of human crural diaphragm electromyogram.
J. Appl. Physiol.
81:
1434-1449,
1996.
|
| 4. |
Beck, J.,
C. Sinderby,
J. Weinberg,
and
A. E. Grassino.
Effects of muscle-to-electrode distance on the human diaphragm electromyogram.
J. Appl. Physiol.
79:
975-985,
1995.
|
| 5. |
Bellemare, F.,
and
A. Grassino.
Effects of pressure and timing of contraction on human diaphragm fatigue.
J. Appl. Physiol.
53:
1190-1195,
1982.
|
| 6. | Cooley, J. W., and J. W. Tukey. An algorithm for the machine calculation of complex Fourier series. Math. Comput. 19: 297-301, 1965. |
| 7. |
Daubenspeck, J. A.,
J. C. Leiter,
J. F. McGovern,
S. L. Knuth,
and
E. J. Kobylarz.
Diaphragmatic electromyography using a multiple electrode array.
J. Appl. Physiol.
67:
1525-1534,
1989.
|
| 8. | Delhez, L. Modalités chez l'homme normal, de la réponse électrique des piliers du diaphragme à la stimulation électrique des nerfs phréniques par des chocs uniques. Arch. Int. Physiol. Biochim. 73: 832-838, 1965. [Medline] |
| 9. | Delhez, L. Electrical responses of the human diaphragm to the electrical stimulation of the phrenic nerves. Electromyogr. Clin. Neurophysiol. 15: 359-372, 1975. [Medline] |
| 10. |
Gandevia, S. C.,
and
D. K. McKenzie.
Human diaphragmatic EMG: changes with lung volume and posture during supramaximal phrenic nerve stimulation.
J. Appl. Physiol.
60:
1420-1428,
1986.
|
| 11. |
Gauthier, A. P.,
S. Verbanck,
M. Estenne,
C. Segebarth,
P. T. Macklem,
and
M. Paiva.
Three-dimensional reconstruction of the in vivo diaphragm shape at three different lung volumes.
J. Appl. Physiol.
76:
495-506,
1994.
|
| 12. |
Grassino, A. E.,
M. D. Goldman,
J. Mead,
and
T. A. Sears.
Mechanics of the human diaphragm during voluntary contractions: statics.
J. Appl. Physiol.
44:
829-839,
1978.
|
| 13. |
Konno, K.,
and
J. Mead.
Measurement of the separate volume changes in rib cage and abdomen during breathing.
J. Appl. Physiol.
22:
407-422,
1967.
|
| 14. | Lindström, L., and R. Magnusson. Interpretation of myoelectric power spectra: a model and its applications. Proc. IEEE 65: 653-662, 1977. |
| 15. |
McKenzie, D. K.,
and
S. C. Gandevia.
Phrenic nerve conduction times and twitch pressures of the human diaphragm.
J. Appl. Physiol.
58:
1496-1504,
1985.
|
| 16. | Mead, J. Discussion. In: Loaded Breathing, edited by L. D. Pengelly, A. S. Rebuck, and E. J. M. Campbell. Don Mills, Ontario: Longman, 1973, p. 206-208. |
| 17. | Merletti, R., M. Knaflitz, and C. J. DeLuca. Electrically evoked myoelectric signals. Crit. Rev. Biomed. Eng. 19: 293-340, 1992. [Medline] |
| 18. |
Milic-Emili, J.,
J. J. Mead,
J. M. Turner,
and
F. M. Glauser.
Improved technique for estimating pleural pressure from esophageal balloons.
J. Appl. Physiol.
19:
207-211,
1964.
|
| 19. | Sarnoff, S. J., L. C. Sarnoff, and J. L. Whittenberger. Electrophrenic respiration. VII. The motor point of the phrenic nerve in relation to external stimulation. Surg. Gynecol. Obstet. 93: 190-196, 1951. |
| 20. |
Schweitzer, T. W.,
J. W. Fitzgerald,
J. A. Bowden,
and
P. Lynn-Davies.
Spectral analysis of human diaphragm electromyogram.
J. Appl. Physiol.
46:
152-165,
1979.
|
| 21. |
Sinderby, C.,
L. Lindström,
and
A. Grassino.
Automatic assessment of electromyogram quality.
J. Appl. Physiol.
79:
1803-1815,
1995.
|
| 22. |
Smith, J.,
and
F. Bellemare.
Effect of lung volume on in vivo contraction characteristics of human diaphragm.
J. Appl. Physiol.
62:
1893-1900,
1987.
|
| 23. | Weinberg, J., C. Sinderby, L. Sullivan, A. Grassino, and L. Lindström. Evaluation of diaphragm electromyogram contamination during progressive inspiratory maneuvers. Electromyogr. Clin. Neurophysiol. In press. |
| 24. | Whitelaw, W. A. Shape and size of the human diaphragm in vivo. J. Appl. Physiol. 55: 1899-1905, 1987. |
This article has been cited by other articles:
|
|
 |
|
  C. Sinderby, J. Beck, J. Spahija, M. de Marchie, J. Lacroix, P. Navalesi, and A. S. Slutsky Inspiratory Muscle Unloading by Neurally Adjusted Ventilatory Assist During Maximal Inspiratory Efforts in Healthy Subjects Chest, March 1, 2007; 131(3): 711 - 717. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. E. Finucane, J. A. Panizza, and B. Singh Efficiency of the normal human diaphragm with hyperinflation J Appl Physiol, October 1, 2005; 99(4): 1402 - 1411. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 ATS/ERS Statement on Respiratory Muscle Testing Am. J. Respir. Crit. Care Med., August 15, 2002; 166(4): 518 - 624. [Full Text] [PDF]
|
|
|
|
|
|
C. SINDERBY, J. SPAHIJA, J. BECK, D. KAMINSKI, S. YAN, N. COMTOIS, and P. SLIWINSKI Diaphragm Activation during Exercise in Chronic Obstructive Pulmonary Disease Am. J. Respir. Crit. Care Med., June 1, 2001; 163(7): 1637 - 1641. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. SINDERBY, J. SPAHIJA, and J. BECK Changes in Respiratory Effort Sensation Over Time Are Linked to the Frequency Content of Diaphragm Electrical Activity Am. J. Respir. Crit. Care Med., March 15, 2001; 163(4): 905 - 910. [Abstract] [Full Text]
|
|
|
|
|
|
J. E. Butler, D. K. McKenzie, and S. C. Gandevia Discharge frequencies of single motor units in human diaphragm and parasternal muscles in lying and standing J Appl Physiol, January 1, 2001; 90(1): 147 - 154. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. M. LUO, R. A. LYALL, M. LOU HARRIS, G. F. RAFFERTY, M. I. POLKEY, and J. MOXHAM Quantification of the Esophageal Diaphragm Electromyogram with Magnetic Phrenic Nerve Stimulation Am. J. Respir. Crit. Care Med., November 1, 1999; 160(5): 1629 - 1634. [Abstract] [Full Text]
|
|
|
|
 |
|
 Y M Luo, M I Polkey, R A Lyall, and J Moxham Effect of brachial plexus co-activation on phrenic nerve conduction time Thorax, September 1, 1999; 54(9): 765 - 770. [Abstract] [Full Text]
|
|
|
|
|
|
J. Beck, C. Sinderby, L. Lindstrom, and A. Grassino Effects of lung volume on diaphragm EMG signal strength during voluntary contractions J Appl Physiol, September 1, 1998; 85(3): 1123 - 1134. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
