| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
1TBMLab, Dipartimento di Bioingegneria, Politecnico di Milano University, I-20133 Milano, Italy; and 2Department of Biomedical Engineering, Boston University, Boston, Massachusetts 02215
Submitted 11 July 2003 ; accepted in final form 9 February 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
lung impedance; chest wall impedance; rib cage; abdomen; forced oscillation technique
Lung impedance (ZL) data at low frequencies has revealed that, during airway provocation akin to asthma or in asthmatic subjects at baseline, airway constriction is highly heterogeneous (22, 28). Also, several related studies suggest that airway reactivity is amplified in the supine condition compared with standing (40), a phenomenon that may have important implications for nocturnal asthma. The implication is that, during supine conditions and/or during provocation, ventilation distribution and chest wall configuration are substantially altered, which also impact the mechanical contributions to breathing. These studies motivate a more thorough assessment of chest wall and total respiratory mechanics in concert with alterations in lung mechanics. In particular, we would like to determine whether local and total changes in respiratory system mechanics reflect heterogeneous alterations in lung function. Recently, a technique called optoelectronic plethysmography (OEP) has been shown to have the ability to measure the volume of the chest wall (Vcw) and to accurately split it into its constituent compartments [i.e., rib cage and abdomen (Ab)] by measuring the position of several passive markers placed on the chest wall surface on selected body landmarks (4, 5, 13, 16).
The main objective of the present study is to combine the OVW method with OEP to study how posture and bronchoconstriction influence mechanical properties of the total respiratory system, the lung, and the chest wall (inclusive of its individual compartments) for frequencies surrounding breathing. Our data were acquired by forcing at the mouth and measuring the flow both at the mouth and at the chest wall surface (as the time derivative of the total and compartmental Vcw) (3, 15). In this way, we can compare input (Zin) and transfer impedances (Ztr) (see Theory) over a frequency range not previously explored by both simultaneously. Thus another objective of this study was to establish if the local impedance and total Ztr (which do not require an esophageal balloon) at lower frequencies are sensitive to important changes in lung mechanics that occur with changes in posture or airway provocation.
Theory.
To first approximation, and particularly at lower frequencies, the respiratory system can be conceptualized as a two-port T-network system (Fig. 1A) in which the airway opening is viewed as one port and the body surface as the second port (38). We define Pao and 
ao as the airway opening pressure and flow, respectively. During forcing at the mouth, ao and body surface flow are different due to the compressibility of alveolar gas and upper airway shunt. When both pressure and flow are measured at the same port of the system, the impedance (defined as Z = P/, where P is pressure and is flow) is called Zin, whereas, if measured at different ports, the impedance is called Ztr. The measurement of pleural pressure [or its estimation by esophageal pressure (Pes)] allows the subdivision of lung from chest wall mechanical properties, as shown in Fig. 1B. From the measurement of the flow displaced by different chest wall compartments, it is possible to consider chest wall models with three or more degrees of freedom. In the present study, we used the three-compartment chest wall model, subdividing the total chest wall into the lung-apposed or pulmonary rib cage (RCp), the diaphragm-apposed or abdominal rib cage (RCa), and the Ab (24). By using this model, it is possible to define several impedances, depending on where flow and pressure are measured (Fig. 1B). In this study, we measured the total Zin, the total Ztr, the lung Zin (ZL), and the chest wall Zin (Zcw). The total Ztr was also subdivided into its compartmental contributions. The definitions of all of these impedances are reported in Table 1.
|
|
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Impedances were measured by asking the subjects to relax passively to functional residual capacity and place their mouth around the OVW mouthpiece while an operator was firmly supporting their cheeks. The ventilator was started such that the first OVW inspiration began from functional residual capacity. The waveform was then presented for 40–100 s. This procedure was repeated three or four times, and the data were stored for subsequent analysis. After this, subjects were administered increasing concentrations of aerosolized methacoline (MCh) solution delivered by a nebulizer system (New Rosenthal Dosimeter, Pulmonary Data Services). Three minutes after each MCh administration, the OVW measurement was performed. We used MCh concentrations of 0.01, 0.1, 1, 10, and 25 mg/ml. The MCh administrations were continued until the maximum dose was reached or until the R at the lowest frequency increased to more than three times the baseline value. We used the lowest frequency R because it is most sensitive to all phenomena associated with bronchoconstriction, including airway and tissue R and heterogeneities. To allow the MCh to wash out, at least 48 h passed before the experiment was repeated in the other posture.
Forced oscillations.
The experimental setup is summarized in Fig. 2. The forcing signal was generated by a personal computer connected to an analog-to-digital-digital-to-analog board (DAQ 6068E, National Instruments, Austin, TX) and sent to the system used to generate the flow forcing. This system was made by a servo-controlled linear motor (Infomag model 7315–1, Goleta, CA) connected to a piston-cylinder arrangement, as described in Ref. 30. The flow-forcing signal was made by adding six sinusoids, whose frequencies were chosen according to the nonsum-nondifference order 3 criteria (41). The sinusoids were chosen with frequencies of 0.156, 0.39, 0.86, 1.48, 2.42, and 4.61 Hz, with the lowest frequency amplitude being three- to fourfold bigger than the amplitude of the other components. The amplitude of the resulting signal was adjusted to provide a tidal volume of 
600–800 ml.
|
ao was measured with a pneumotachograph (Fleisch no. 2) connected to a pressure transducer (model TD05, ±5 cmH2O; SCIREQ, Montreal, QC). Pes was measured by a pressure transducer (model SCX05, SenSym, Milpitas, CA) connected to a 10-cm-long thin latex balloon positioned in the lower one-third of the esophagus and filled with 0.4 ml of air. Transpulmonary pressure was obtained by subtracting Pes from Pao. The position of the esophageal balloon was considered correct if 1) transpulmonary pressure remained almost unchanged during gentle efforts against a shutter connected to the mouthpiece and closed at end expiration of a tidal breath (10), and 2) if the cardiogenic oscillations were minimized. Special care was used for the balloon positioning in the supine position (11). All of the signals from the transducers were amplified, low-pass filtered at 40 Hz (model SC-24, SCIREQ), and digitized at 200 Hz by the same analog-to-digital-digital-to-analog board used to generate the forcing signal. Vcw was measured plethysmographically by OEP, a recently introduced technique extensively validated in several conditions and postures (2, 4, 5, 13, 17, 24). Briefly, several (89 for the seated position and 68 for the supine) reflective markers were placed over the thorax from the clavicles to the anterior superior iliac spines along predefined vertical and horizontal lines. The position of the markers is described in Ref. 13 for the seated position and in Ref. 15 for the supine. The three-dimensional position of each marker was measured by an automatic motion analyzer (ELITE System, BTS, Milano, Italy) at a frequency of 100 Hz by using four special video cameras (two in front and two behind the subject for the seated position, two on the left-hand side and two on the right-hand side above the subject for the supine, Fig. 2). Vcw was finally determined by approximating chest wall surface by triangles connecting the markers and computing the volume enclosed by all of these triangles. The use of a subset of the markers allowed us to measure separately the volume of the three chest wall compartments defined in the Theory section, subdividing Vcw into the RCp volume, the RCa volume, and the abdominal volume. These data permit estimation of local impedances (Table 1). This technique was proven to be reliable when used to measure total and compartmental tidal volume, as well as to estimate the total respiratory system Ztr, at least up to 24 Hz (3). Moreover, because it is based on the measurement of volumes, it is not affected by integration drift, and, therefore, it can be used to track breath-by-breath end-expiratory lung volume changes (16) and thus measure the possible total and compartmental dynamic hyperinflation induced by bronchoconstriction.
Data analysis.
Impedances Z have been estimated as follows (32):
 | (1) |
The estimates of Gpf and Gff have been obtained by averaging different periodograms computed by applying the fast Fourier transform to several data segments. To reduce the picket-fence effect, we resampled the acquired data at a new sampling frequency of 40 Hz. The resampling procedure was preceded by an anti-aliasing low-pass digital filter compensated for the filter's delay. Because our forcing signal period was 12.8 s, one OVW period after the resampling was 512 samples long. In this way, all frequencies of the OVW signal perfectly matched the discrete frequency points for which the fast Fourier transform computes the exact behavior.
The signal segments were selected manually by an operator to discard those with esophageal spasms or breathing efforts. The Ztr were computed by measuring pressure and volume and by multiplying the Fourier transform of the volume signal by j
, where j is the imaginary unit, is equal to 2
f, and f is the frequency (in Hz). This procedure is equivalent to differentiating the volume signal in the time domain to obtain the flow signal (3).
The coherence functions of the impedances were computed as described in Ref. 3. In the present study, we considered only values of impedances with a coherence value >0.90.
The transfer functions of the entire population of subjects were finally obtained by averaging the real and the imaginary part of the transfer function values from all of the subjects at all of the frequencies. If the value of the impedance was discarded for one frequency because its coherence was below the threshold (likely due to cardiogenic oscillations in Pes or to low signal-to-noise ratio in volume signal), it was estimated by linear interpolation by using the adjacent values.
In this study, all of the impedances are presented in terms of R, i.e., the real part, and E, which were obtained by multiplying the imaginary part by –2f, where f is the frequency in Hz.
Statistical analysis. Values are presented as means ± SE, if not otherwise specified. A nonparametric Wilcoxon test determined statistical significance of changes in all of the considered transfer functions with changes in posture or after bronchoconstriction. The test was performed for both R and E at each forcing frequency. Significance was taken as P < 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
|
To further examine the impact of posture change, we can partition Ztr into its local impedance components (Fig. 5). Here we do not show the RCa compartment, as its contribution to Vcw was found to be negligible compared with the RCp and Ab. A large increase in the RCp R can be observed when passing from the seated to the supine position. Conversely, the R of the abdominal compartment showed a small reduction in the supine position compared with the seated one. Therefore, while in the seated position the two compartments showed comparable R values, in the supine position they offered very different R values. Due to the fact that the two compartments are in parallel (they are submitted to the same pressure, Pes), and because in the supine position the RCp impedance had a very high baseline value, in this posture the total Ztr R is determined mostly by the abdominal compartment, showing only small differences compared with the seated posture.
|
Ztr vs. ZL during bronchoconstriction. In Fig. 6, the effects of bronchoconstriction on ZL, Zcw, and Ztr are shown for the supine position. We show supine only because the effect of the MCh was greatest in this position. Most of the effects of MCh on Ztr are due to changes in lung mechanics. In fact, ZL dramatically increased both in mean level and in frequency dependence. As reported previously (26), these changes are highly consistent with acute heterogeneous constriction occurring throughout the peripheral airway tree, inclusive of airway closures or near closures. Hence the noninvasive measure of Ztr via the OEP method is highly sensitive to heterogeneous constriction in the lung.
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
However, no previous method could acquire Ztr reliably at low frequencies in human subjects. This is because the measurement of Ztr requires either the linear excitation of pressure surrounding the entire chest wall or a direct measure of chest wall flow. The former is not practical, especially with loudspeaker technology.
In this study, we defined a new method to measure low-frequency mechanical properties of the respiratory system by combining the optimal ventilation waveform technique, used to provide both ventilation and forcing signal, with the OEP, used to measure chest wall flow. There are several advantages to this approach of measuring Ztr: 1) it does not require the use of a plethysmographic chamber, and it can be applied in different postures and conditions; 2) it allows the separation between the different rib cage and abdominal pathways; 3) it allows the simultaneous assessment of Zin and Ztr; and 4) it allows the breath-by-breath tracking of lung volume changes and chest wall configuration through the direct measurement of total and compartmental end-expiratory Vcw. Moreover, if combined with Pes measurement, this technique allows the partitioning of lung and chest wall mechanical properties by the measurement of the impedance of each single component (see Theory).
Comparison between Zin and Ztr. The first important step was to validate that the OEP-based method provided a reliable approach to estimating Ztr. Previous studies by our group have validated the ability of OEP to measure volume changes and changes in end-expiratory lung volumes (4, 13), as well as its ability to measure Ztr during single sinusoidal forcing at higher frequencies (4–20 Hz) (3, 15). Validation of OEP at the lower frequencies used in this study is evident from two aspects of our data. First, despite the fact that ZL was measured by pneumotachography whereas Zcw was measured by OEP, the sum of ZL and Zcw was very close to Ztr (see RESULTS). Second, several other studies have presented data on low-frequency Zin (6, 20, 30), and they reported Zin values that are similar to ours. To our knowledge, there are no other studies presenting low-frequency Ztr measured by forcing at the mouth. However, we expected that Zin and Ztr would differ only because of a distinct influence of shunt impedances on Zin. Specifically, when forcing at the mouth, upper airway shunt and alveolar gas compression will tend to reduce Zin but not impact Ztr (27, 36). This effect should be least evident during the control sitting conditions and more evident as ZL increases. Indeed, we found that Zin and Ztr were very similar for control and seated position measurements and that moving from seated to supine caused the R of Ztr to increase slightly compared with that of Zin (Fig. 4). After bronchoconstriction, the differences between Zin and Ztr became more substantial (Fig. 7), and this is consistent with an increased influence of shunt on Zin but not on Ztr.
In this study, we partitioned chest wall from lung mechanical properties by estimating pleural pressure swings by the esophageal balloon technique in both the seated and the supine positions. It has been shown that, in the supine posture, Pes swings represent an accurate estimation of pleural pressure swings (11). This is supported by the very similar values obtained for low-frequency lung E in the two postures (6.1 ± 0.8 cmH2O/l for the seated position and 6.7 ± 0.7 cmH2O/l for the supine).
In summary, Ztr and Zcw measured by using OEP are consistent with Zin and ZL simultaneously measured by pneumotachometry, showing that the combination of OEP and OVW provides a reliable method for measuring low-frequency impedances.
Analysis of respiratory impedances and the impact of posture.
In the seated position, lung R (Fig. 4) is consistent with a relatively homogeneous mechanical system with some slight frequency dependence due to viscoelastic lung tissues (23). Most of the frequency dependence of Ztr at low frequency is, therefore, due to the chest wall tissues. Our data show that, around breathing frequencies, the chest wall contributes 62% to the total Ztr R in the seated position, whereas, at 4.6 Hz, Zcw contributes only 24% of the total Ztr. Similar results for Zcw were found by Barnas et al. (8), measuring chest wall flow by a partial body plethysmograph, and by Hantos et al. (20), estimating Zcw from Zin after correcting for upper airways and thoracic gas shunt.
The change in posture slightly increased the mean level and frequency dependence of ZL (Fig. 4). These changes are compatible with the differences in lung volumes and airway diameters anticipated for supine vs. seated positions. The increase in frequency dependence of R and E is consistent with the supine position, enhancing airway heterogeneities even at baseline. A reduced lung volume causes reduction of the elastic recoil forces acting on the airways, which is combined with the change of the dependent regions of the lung (due to the different direction of the gravitational forces), inducing a different degree of airway constriction throughout the bronchial tree.
Very little is known about the effect of posture on Zcw, likely due to the difficulties in measuring chest wall flow in the supine position. In the supine position, we found that both Zin and Ztr presented a substantial increase in their low-frequency components (Figs. 4 and 7). Interestingly, even if Ztr R at the lowest frequency increased from 6.9 to 10.8 cmH2O·l–1·s passing to the supine position, the relative contribution of the Zcw to Ztr was almost the same as in the seated position at the lowest frequency (63%), whereas it slightly increased at the highest frequency (36%). In 1993, Barnas et al. (7) measured the effect of posture on Zcw by using a sinusoidal forcing at 0.2 Hz. They reported a baseline Zcw value and changes in Zcw due to posture that were smaller than ours. However, they used ao to compute Zcw instead of chest wall flow, and, therefore, the effect of upper airway and cheek shunting might be responsible for these differences.
Compartmental analysis. The OEP technique allowed us to partition the volume changes of chest wall into compartments. Such partitioning allowed us to subdivide Ztr into its compartmental pathways. Implicit in the definitions of the compartmental impedances of Table 1 is that all of the pathways are independent. However, as Loring and Mead (25) first pointed out, these compartments are not mechanically independent, as RCa and Ab are submitted to the same pressure (abdominal pressure) and RCa and RCp are mechanically linked by anatomic structures (the ribs and the sternum). Several models have been suggested that include the effect of the different pressures acting on the rib cage as well as the transmission of forces through mechanically coupled structures (12, 19, 21, 31, 43). In our study, the measurements occur with relaxed chest wall. Hence, we decided to ignore the coupling between compartments, as its effect on impedance estimation should be negligible.
Figure 5 shows that, in the seated position, the RCp and the Ab displayed similar R and E values, making the whole chest wall behave like a single, homogeneous structure. The results in the supine position are completely different, with the RCp almost doubling both its R and E and the Ab slightly reducing them. Because the two pathways are in parallel, the total Ztr in the supine position displays only a small increase in R and essentially the same E as in the seated position. Nevertheless, the chest wall behaves more inhomogeneously in the supine position, with the Ab being the most important compartment in determining total respiratory system mechanics. Therefore, in the supine position, changes in the abdominal mechanical properties have a much bigger effect on the total respiratory mechanics than in the seated position. This may have important implications regarding patient care in the intensive care unit. Such patients may experience abdominal distension from various origins (39) (as, for example, surgical acute respiratory distress syndrome patients) (35), which increases abdominal stiffness (33). Based on our results, this condition will greatly amplify the Ztr and Zin in the supine position, which will make spontaneous breathing and mechanical ventilation far more arduous.
This increased influence of the Ab in the supine position is likely due to the direction of the gravitational forces relative to the physical support of the thorax. Specifically, in the supine position, most of the weight of the thorax is supported from the back of the subject, with the ribs supported directly by the bed and, therefore, less free to move and expand the rib cage. Conversely, the Ab has no rigid structure around the back (excluding the spinal chord), and the most important effect changing to the supine position is the change of direction of gravity on the abdominal contents. Thus, in supine, the abdominal weight is no longer stretching the abdominal walls, and it is now pushing on the back (where it is supported by the bed). The result is that the abdominal contents are moved cranially, pushing the diaphragm up and lowering the lung volume (1). In this condition, the Ab has more freedom to move, as shown by its lower R and E.
In summary, under the hypothesis of a uniformity of pleural swings around the lung, the combination of OEP and OVW allows us to noninvasively track alterations in regional chest wall mechanics, providing potentially useful information for patients with restrictive disease, those in the intensive care unit, and for postoperative patients, especially those undergoing upper abdominal surgery.
Effect of bronchoconstriction. Bronchoconstriction had two major effects on the respiratory system. The most important one was on the ZL, where heterogeneous constriction of the airways has been shown to induce an increase of both mean level and frequency dependence of R. The second effect is on the chest wall configuration, with dynamic hyperinflation causing the end-expiratory Vcw to increase differently in each compartment and thus changing the range of the compartmental pressure-volume curve over which the subject is being ventilated.
As shown in Fig. 6, after MCh, lung R and E markedly increased with minor effects on Zcw. The absence of changes of Zcw indicates that, even if the subjects hyperinflated differently between the compartments when constricted (Table 2), this change was not sufficient to induce significant changes on the total mechanical properties of the chest wall. Figure 6 also shows that, because Zcw did not change significantly, the impact of bronchoconstriction on ZL is accurately reported by the changes of Ztr, as it is the series of the two respiratory structures. Thus the combination of OVW and OEP also provides a noninvasive tool for measuring the impact of bronchoconstriction on the lung without requiring the esophageal balloon. Conversely, the effect of bronchoconstriction is underestimated by Zin, as shown in Fig. 7, where the real part of Zin lies below Ztr, particularly in the supine position where the level of constriction is higher. The amplified sensitivity of Ztr to MCh compared with Zin in the supine position is consistent with a greater effect of airway wall shunting on Zin during the challenge.
It is interesting to compare how the compartmental pathway impedances are modified by MCh in the seated vs. the supine positions. In Fig. 8, the ratios between the real part of the RCp and the abdominal pathways are presented in the two postures before and after MCh. Values greater than one indicate that the RCp presents a larger R than the Ab. The higher the value, the greater the RCp R and, therefore, the greater the amount of in-phase flow displaced by the Ab for a given Pao. Bronchoconstriction induced different effects on the two postures. In the seated position, it markedly changed the frequency dependence of the compartmental contributions. Here, at high frequency, the RCp R was 0.8 times the abdominal R, both before and after MCh, whereas, at low frequency, it was 1.1 times in control conditions and to 1.5 times after MCh. In the supine position, the contributions of the two compartments presented smaller frequency dependence than in the seated position, with the RCp R being 2.5 times the abdominal one at the lowest frequency and 1.7 at the highest frequency. After bronchoconstriction, even if the level of constriction reached in the supine position was higher than in the seated one, the relative contribution of the RCp was 2.5 at the lowest frequency and 1.3 at the highest, suggesting that posture affects the relative changes of compartmental mechanical properties differently. It is still unclear whether the differences observed are the results of different mechanical properties of the chest wall compartments or whether the heterogeneity of bronchoconstriction throughout the lung has a measurable effect on macroscopic regional mechanics, resulting in regional heterogeneity of pleural pressure swings.
|
|
GRANTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Meinero, G. Coletta, L. Dutto, M. Milanese, G. Nova, A. Sciolla, R. Pellegrino, and V. Brusasco Mechanical response to methacholine and deep inspiration in supine men J Appl Physiol, January 1, 2007; 102(1): 269 - 275. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Torchio, C. Gulotta, C. Ciacco, A. Perboni, M. Guglielmo, F. Crosa, M. Zerbini, V. Brusasco, R. E. Hyatt, and R. Pellegrino Effects of chest wall strapping on mechanical response to methacholine in humans J Appl Physiol, August 1, 2006; 101(2): 430 - 438. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
), chest wall impedance (Zcw; 
), Zin (
), and Ztr (
) in control conditions. Data are presented for the seated (left) and the supine (right) positions. Values are means ± SE. *P < 0.05, seated vs. supine.
) and Ztr (
) of the total respiratory system expressed as resistance (top) and elastance (bottom) in the seated (left) and the supine (right) positions before (solid lines) and after (dotted lines) bronchoconstriction. Values are means ± SE. *P < 0.05, control vs. MCh.