HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Free All Versions of this Article: 104/4/1177    most recent 00212.2007v1 Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (2) Citing Articles via Google Scholar Google Scholar Articles by Fuld, M. K. Articles by Simon, B. A. Search for Related Content PubMed PubMed Citation Articles by Fuld, M. K. Articles by Simon, B. A.

CT-measured regional specific volume change reflects regional ventilation in supine sheep

Departments of ¹Anesthesiology and Critical Care Medicine, and ²Medicine, The Johns Hopkins University, Baltimore, Maryland; and Departments of ³Radiology, and ⁴Biomedical Engineering, University of Iowa College of Medicine, Iowa City, Iowa

Submitted 20 February 2007 ; accepted in final form 4 February 2008

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
Computer tomography (CT) imaging techniques permit the noninvasive measurement of regional lung function. Regional specific volume change (sVol), determined from the change in lung density over a tidal breath, should correlate with regional ventilation and regional lung expansion measured with other techniques. sVol was validated against xenon (Xe)-CT-specific ventilation (s) in four anesthetized, intubated, mechanically ventilated sheep. Xe-CT used expiratory gated axial scanning during the washin and washout of 55% Xe. sVol was measured from the tidal changes in tissue density (H, houndsfield units) of lung regions using the relationship sVol = [1,000(Hi – He)]/[He(1,000 + Hi)], where He and Hi are expiratory and inspiratory regional density. Distinct anatomical markings were used to define corresponding lung regions of interest between inspiratory, expiratory, and Xe-CT images, with an average region of interest size of 1.6 ± 0.7 ml. In addition, sVol was compared with regional volume changes measured directly from the positions of implanted metal markers in an additional animal. A linear relationship between sVol and s was demonstrated over a wide range of regional s found in the normal supine lung, with an overall correlation coefficient (R²) of 0.66. There was a tight correlation (R² = 0.97) between marker-measured volume changes and sVol. Regional sVol, which involves significantly reduced exposure to radiation and Xe gas compared with the Xe-CT method, represents a safe and efficient surrogate for measuring regional ventilation in experimental studies and patients.

multidetector-row computer tomography scanning; xenon computer tomography; implanted markers; lung mechanics

As lung volume increases, overall CT density falls due to the inflow of gas. This CT density is linearly related to the F of the tissue. If we assume that 1) all volume changes are due to the increase in air volume and 2) the volume distributes uniformly throughout the region, then we can estimate change in volume by measuring the change in air content of a lung region at two pressures. By normalizing change in volume by the initial regional air volume to give sVol, the volume terms are eliminated from the equation and determination of sVol requires only measurement of the regional density at the two volumes. Furthermore, since the regional volume terms do not appear in the final expressions and since lung density changes relatively smoothly over a scale greater than a few millimeters, it is not necessary to determine the exact same tissue elements in both regions of interest (ROIs) but only to roughly encompass the same region (which can be easily identified by specific anatomical markings).

The Xe-CT method for measuring regional ventilation relies on the determination of the time constant of regional lung density changes from serial CT scans taken at end expiration during the multi-breath washin and/or washout of the radiodense tracer gas Xe (9, 18, 28). Although this approach permits the measurement of regional ventilation with very fine spatial resolution and a high degree of anatomical localization, it involves repeated imaging at each axial location with the associated radiation exposure. In addition, Xe gas is expensive, requires specialized equipment for delivery, and has anesthetic and sedative properties that limit the concentration, and hence the maximal contrast enhancement and signal-to-noise ratio, at which it may be used in humans (15, 32).

Implanted radiodense parenchymal markers, tracked using X-ray techniques, have been used to quantify regional lung motion and parenchymal strain in animal models and address fundamental issues in regional lung mechanics in healthy and injured animal models (12, 19). Since the volume of lung as well as the mean CT density contained within a tetrahedral volume element (TET) defined by four markers can be calculated, we utilized a set of CT images from a pilot study in sheep to compare directly measured regional lung volume changes to that predicted from the density change using our sVol formula.

The rationale for using sVol instead of a simple straight change in regional density is that, as we will illustrate with a mathematical model, the density change is nonlinearly related to ventilation and is highly dependent on both the initial region density and the magnitude of the density change. This new method eliminates much of the exposure and cost associated with techniques using ventilation tracer gases. It allows for the use of traditional CT techniques, coupled with respiratory gating, for a simpler and more efficient measure of regional ventilation.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
Animal preparation.   All animal studies were approved by both the Johns Hopkins University and University of Iowa Animal Care and Use Committees. Four farm-bred sheep were premedicated with 4 mg/kg ketamine intramuscularly and anesthetized with 5% halothane by nose cone inhalation. Femoral arterial and external jugular venous lines were placed, and a tracheostomy was performed with a 9.0-mm inner diameter cuffed endotracheal tube. Anesthesia was maintained by intravenous pentobarbital (1–3 mg·kg^–1·h^–1 and as indicated), and the animals were relaxed with pancuronium (0.1 mg/kg iv and 0.5–1 mg hourly). Arterial pressure, oxygen saturation, and airway pressures were continuously monitored and recorded. Animals were placed in the supine position and were held with gentle forelimb traction. Mechanical ventilation with 100% oxygen, tidal volume of 10 ml/kg, rate of 20 breaths/min adjusted to end-tidal PCO₂ of 30–35 Torr, and 5-cmH₂O positive end-expiratory pressure was applied. A computer-based system for monitoring physiological and experimental parameters was implemented using an IBM-compatible computer equipped with a standard laboratory analog-to-digital converter board and a customized LabVIEW 7.0 (National Instruments, Austin, TX) interface. The software was programmed to record physiological information, including airway pressure, end-tidal PCO₂, airway opening Xe concentration, arterial pressure, and ECG, which were then correlated with CT scanner events to allow gating the pulse of the scanner X-ray to specific points in the ventilation cycle. During a Xe-CT scan, radiodense Xe gas was introduced into the breathing circuit from a concentration-controlled reservoir (Enhancer 3000, Diversified Diagnostic Products, Houston, TX). The reservoir was attached to a custom dual-Harvard piston-ventilator setup (Harvard Apparatus, Holliston, MA) that was specifically designed for external computer control, allowing for remote switching between room air and an O₂-Xe mixture. Delivered tidal volumes were matched between both ventilators using a calibrated pneumotachograph (model 3700A, Hans Rudolph, Kansas City, MO).

CT imaging.   Imaging was performed on a Marconi MX8000 four-slice scanner, modified for external gating of image collection. Ventilation images were obtained using the washin-washout Xe-CT method (26), in which a series of axial images were obtained at a fixed table position gated to both end-expiration and end-inspiration for 70 breaths. Each series included 10 baseline (100% O₂), 30 Xe washin (55% Xe, 45% O₂), and 30 Xe washout images (100% O₂). Only the end-expiratory images from the series were analyzed to determine regional ventilation. The density change over time in each lung parenchymal voxel in the series was determined and fit to an exponential, the time constant of which is equal to the inverse of the voxel-specific ventilation, as previously described (26). Scanner settings for ventilation imaging were slice thickness of 2.5 mm (4 slices simultaneously obtained each image acquisition), 100 kV, 90 mA, angle of 360, and scan time of 0.53 s. Repeat Xe studies were performed after incremental table movement in each animal to obtain 8–12 contiguous 2.5-mm-thick slices (2–3 cm total axial coverage) in each of the apex and base regions. Inspiratory and expiratory images used for the sVol calculation were obtained from the baseline images in the Xe-CT scan before Xe inhalation, thus ensuring that the tidal excursions from both analyses were the same. Images were segmented to separate lung tissue from chest wall and mediastinum using the Pulmonary Analysis Software Suite (PASS) and then further analyzed as described below using the National Institutes of Health's Image J software (1). Ventilation images were analyzed using the time-series image analysis" (TSIA) software. PASS and TSIA were developed by the University of Iowa, Division of Physiological Imaging (11).

Calculating sVol from CT air content.   The relationship between regional density change and regional volume change may be derived using a mass balance approach. A lung ROI is defined as an identified region of lung tissue whose boundaries move as it inflates and deflates. We then define V_e as the total volume of a given ROI_i at end-expiration and V_i as its total volume at end-inspiration. Similarly, V_ae and V_ai are the air volumes of the ROI and F_e and F_i are the ROI Fs at end-expiration and end-inspiration, respectively. We make the following assumptions: 1) a pressure increase results in a volume increase (V) to a ROI, 2) this volume change is due to an increase in air volume and results in a rise in the F, and 3) density is uniform within the ROI. The ROI air and total volumes are related by the Fs:

(1)

(2)

Volume change:

(3)

But since we assume that all volume change is due to the air volume change:

(4)

Substituting in Eqs. 1 and 2

(5)

Solving for V_i yields:

(6)

sVol is defined as the change in volume normalized by the initial air volume:

(7)

To express sVol in terms of of the measurements of air content F, Eqs. 1, 3, and 5 were combined and Eq. 6 was substituted to obtain:

(8)

Note that volume does not appear in the final formula, only density. Since density is a comparatively smooth function of position in the lung parenchyma, the result should be less sensitive to small registration errors than an approach that seeks to manually measure specific lung ROI dimensions as they change volume across conditions (see DISCUSSION).

Finally, using the idealized relationship between CT density (Hounsfield Units, HU) and F (H = –1,000F, where H is mean CT density in the ROI in HU), we can substitute into Eq. 7 to get an equation utilizing density measurements in HU:

(9)

[Note that this assumes air and pure tissue densities of –1,000 and 0 HU, respectively, which may not be accurate for a specific scanner. For scanners with known or measured air and tissue densities of H_A and H_T, HU, the relationship H = H_T – (H_T – H_A)F should be used.] Thus regional lung sVol may be measured from the changes in tissue density of lung regions that occur with changes in inflation pressure by comparing regions imaged at end-expiration and end-inspiration using Eq. 9 above.

Finally, it should be noted that by using end-expiratory (EE) and end-inspiratory (EI) images a measure of regional specific volume change is obtained that is closely related to tidal ventilation. However, any images of the lung at two lung volumes, such as functional residual capacity and total lung capacity, can be substituted, and the resultant values will reflect the regional volume change over that pressure and volume excursion. Furthermore, one can obtain regional specific compliance if airway or transpulmonary pressures are available by simply dividing sVol by the change in pressure. In this case, the regional specific compliance, or compliance per unit of air volume (cmH₂O^–1), will reflect respiratory system mechanics if airway pressure is used and lung properties if transpulmonary pressure is used.

Xe-CT ventilation method.   Since stable Xe gas is denser than air, its CT enhancement is linearly related to its concentration (26). The density values for a ROI from a series of images taken at EE during the washin and/or washout of 30–70% Xe will yield a curve that can be fitted to a single exponential model with a time constant . This time constant is equal to the inverse of the specific ventilation (s), the ventilation per unit volume: = 1/s (26).

Image analysis.   Distinct anatomical markings in the images, such as airway or vessel branch points, were used to match the same lung ROIs between baseline EE, baseline EI, and Xe-CT (EE) images (Fig. 1). That is, we used these distinctive markings to choose the CT slices that best matched identifiable anatomical features and landmarks between the lung volumes, since the lung can move in relation to the CT table (which defines the slice number or Z position) with respiration and over time. The average ROI size used was 1.6 ± 0.7 ml (mean ± SD). Changes in ROI density with tidal breathing were measured with the public domain program Image J (1) and used to calculate sVol. Corresponding regional Xe-CT s was measured with TSIA software from the University of Iowa (4, 27) (Fig. 1A), which uses a nonlinear curve-fitting routine to determine the regional time constant that simultaneously fits the washin and washout data (26). Data for s were rate normalized to account for varying respiratory rates (RR) between animals, yielding a measurement of ventilation per breath that is appropriate for comparison to the sVol parameter.

(10)

View larger version (76K): [in this window] [in a new window]   Fig. 1. Examples of manual region of interest (ROI) matching between the Xe-computer tomography (CT) ventilation study (A), using the University of Iowa's TSIA software, and the gated expiratory and inspiratory images (NIH Image J; B and C) used to determine specific volume change (sVol). The three curves of A represent the time-radiodensity curves from the Xe-CT study for the colored ROI indicated.

Parenchymal marker implantation and imaging.   An additional anesthetized, mechanically ventilated sheep was studied using a different technique. Sixty-nine 1-mm stainless steel beads were implanted throughout the lung parenchyma of the left apex (34 beads) and right base (35 beads) during open surgery, and then the animal was allowed to recover for 6 wk before further analysis. Whole lung volume CT scanning of the anesthetized, intubated, mechanically ventilated animal in the supine and prone positions at constant airway pressure of 0, 8, 16, 24, and 32 cmH₂O was performed.

Implanted marker analysis.   After recovery, the steel beads throughout the lung parenchyma were used to track movement and expansion of lung parenchyma during breathing. Three-dimensional marker positions were determined and tracked across different airway pressures via an image registration process (7, 16, 21). Individual markers were grouped into TETs, screened for size and shape criteria (minimum volume of >0.5 ml and aspect ratio >0.35) and the volume, CT density, and location of each TET determined for each condition. These TETs allowed for a direct measurement of lung expansion at each level of inflation. Since TET volume and mean density were each measured, the normalized TET volume changes can be directly compared with the measurement of sVol based on density (Eq. 9). The following formula was used to characterize sVol*, the change in TET volume normalized to starting air volume for appropriate comparison to sVol:

(11)

where V is TET volume, H is TET mean density in HU, and indexes 1 and 2 indicate starting and ending volume for each step. A total of 541 tetrahedra were tracked across all four pressure steps.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
An examination of data from multiple animals revealed a linear relationship between sVol and s (note that regressions were calculated assuming that zero sVol will equal zero s). For ROI within a single slice, R² values approached 1 and were routinely above 0.80 (Fig. 2). Within an individual animal, including multiple slices at the apex and base, the correlations ranged from 0.56 to 0.76 (Fig. 3). The average value of R² across the apex regions was 0.81 (range 0.72–0.88) and across basal regions was 0.71 (range 0.63–0.76; Fig. 4), and the slopes were not significantly different (P = 0.2). The relationship between sVol and s held over a range of ROI sizes, with greater scatter in the data from smaller ROIs yielding the following correlation coefficients: small R² = 0.59, medium R² = 0.71, large R² = 0.76 (Fig. 5A). All correlation coefficients were statistically significant with P < 0.002. There were similar vertical gradients in regional s and sVol, with more gravitationally dependent regions receiving greater relative ventilation and volume changes (Fig. 5B), as has been previously described (9, 18, 28). The comparison of sVol from density changes and sVol* from the implanted markers is presented in a Bland-Altman plot (3) demonstrating low bias but relative errors (difference divided by the mean) approaching 50% at the lowest sVol values (Fig. 6). Svol and sVol* were very strongly correlated, with linear regression yielding the relationship sVol* = 1.01 x sVol – 0.0016 (R² = 0.96, P < 0.0001). Note that the largest sVol occur over the lower pressure ranges, as expected from the higher lung compliance at lower lung volumes.

View larger version (12K): [in this window] [in a new window]   Fig. 2. Relationship between sVol and Xe-CT-specific ventilation (s) for multiple ROI (average 1.5 ml ROI volume) taken from a single apical slice.

View larger version (24K): [in this window] [in a new window]   Fig. 3. Composite sVol and Xe-CT s data for apex and base regions in four animals with an average ROI size of 1.6 ± 0.7 ml. There was a combined correlation coefficient R2 of 0.66 (range 0.56–0.76).

View larger version (19K): [in this window] [in a new window]   Fig. 4. Relationship between sVol and Xe-CT s stratified by apex vs. base location.

View larger version (77K): [in this window] [in a new window]   Fig. 5. Effect of ROI size and vertical position on sVol in a single slice. A: the increased scatter at the smallest ROI size (small R2 = 0.59, medium R2 = 0.70, large R2 = 0.76). B: similar vertical gradients in s and sVol, with increased values in ROI with more gravitationally dependent locations at all ROI sizes.

View larger version (21K): [in this window] [in a new window]   Fig. 6. Bland Altman plot of sVol measured from implanted marker displacements (sVol*) and from CT density changes (sVol) for 541 tetrahedral volume elements as airway pressure increased from 0 to 32 cmH2O in four steps. Broken line indicates mean, and solid lines indicate the 95% confidence limits for agreement (10). Note that, although the mean bias is small, the relative error at small sVol approaches 50% for some tetrahedra.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

As the lung inflates and fills with air, its density falls. However, although the fall in CT density of a lung ROI is related to the regional ventilation or volume change, the relationship is nonlinear and dependent on both the magnitude of the density change and the initial region density. This phenomenon can be modeled using the same mass balance used to derive the sVol relationship above, assuming that all lung tissue elements are conserved, that density falls solely due to the influx of air and that air and tissue have densities of –1,000 and 0 HU, respectively. This model predicts that the sVol, or volume change/initial air volume, for a given change in ROI density can vary more than threefold as the initial region density changes from –500 to –850 HU, values of lung density that may be found in a normal lung under different conditions (Fig. 7). As a result, equating straight CT density change with regional ventilation (6, 30) is likely to introduce error. In contrast, the sVol relationship derived above (Eq. 8 or 9) corrects for this nonlinearity and has been shown to correlate with whole lung measures of tidal ventilation (7) and Xe-CT ventilation on the level of the axial slice (24).

View larger version (18K): [in this window] [in a new window]   Fig. 7. Model results illustrating the theoretical effect of initial ROI density on sVol for a range of lung density changes ( HU) encountered during normal tidal breathing. The same measured change in density can be associated with a greater than threefold increase in sVol as initial ROI density falls from –500 to –850 HU (more aerated lungs). HU, houndsfield units.

View larger version (107K): [in this window] [in a new window]   Fig. 8. Effect of small registration errors in ROI identification on ROI volume and mean density. Although the volumes of these three closely overlapping ROI vary 10–20%, the mean density varies by <0.2%.

Selecting the ideal ROI size to minimize scatter in the data is important. An examination of sVol vs. s in a single CT slice over a variety of ROI sizes yields similar linear relationships with partial volume and noise effects reduced significantly, moving from small to medium ROI size (Fig. 5A). For smaller ROI sizes, there was more scatter in the data, and the correlation R² values increased from 0.59 for small regions to 0.76 for large regions obtained from the same slice (Fig. 4). Similarly, Guerrero et al. found a high degree of correlation between sVol summed for the entire lung and overall volume changes in human inspiratory and expiratory CT images obtained for radiotherapy treatment planning (7).

There are several potential reasons for this decreasing correlation with ROI size and location. Noise may increase as ROI size gets smaller in both the sVol and s measurements. First, partial volume effects and registration errors may introduce noise in the individual ROI density measurements. We manually matched ROI between EE and EI in single CT slices using clearly identifiable anatomical landmarks and used 2.5-mm-thick slices as a tradeoff between more robust ROI averaged density measurements in thicker slices vs. reduced in-plane registration error with thin slices. However, the lung displaces caudally with inspiration, and, particularly toward the base, this displacement can be nonuniform between dependent and nondependent regions and thus may not register precisely in-plane. Although, as discussed above, the use of average ROI density should be less sensitive to registration errors compared with tracking ROI volume, partial volume effects of dense structures such as airways or vessels may contribute to sVol noise, and this effect would be expected to decrease with increasing ROI size. The use of three-dimensional registration techniques for ROI matching are under development and should improve these problems in the future (7, 16, 21). Finally, the increasing noise seen in the data with decreasing ROI size may also reflect the scale-dependent heterogeneity of ventilation and perfusion in the lung, in which the heterogeneity of ventilation and perfusion are observed to increase as the measurement scale decreases (22). Ultimately, whether this relationship between noise and ROI size will limit the utility of this method will depend on the question being addressed, but for many applications it can provide a reasonable measure of regional ventilation down to a scale of 1–2 cm.

Noise from the Xe-CT measurement likely also contributes to the observed scatter. We used a traditional model in which washin and washout Xe time constants were constrained to be equal (26), whereas recent results suggest that these time constants may differ and that the differences vary with location in the lung, ventilator settings, and the concentration of Xe gas used (4). Since this is a multi-breath imaging study, misregistration of images as the lung returns to its end-expiratory volume after each breath is a major source of noise, an effect that would also increase at smaller ROI volumes due to partial volume effects. The density of inhaled Xe may also affect its distribution in the lung compared with air. Previous studies have shown that the 95% confidence interval for Xe-CT s measures gets larger as ROI size is reduced (26). Cardiac motion contributes a large component of noise to the Xe-CT images, and this noise is most prominent in regions adjacent to the heart and in the lung base (23, 26).

The implanted marker data compare measured static regional lung volume changes to those predicted from density changes by the sVol formula, perhaps a more appropriate comparison of structural volume changes than to the functional Xe-CT gas transport data. These data show a tight quantitative correlation over a wide range of lung volumes and pressure steps, providing strong direct validation of this approach.

It is important to reiterate several significant limitations to this study. As mentioned above, the quantities sVol and regional ventilation measure fundamentally different things (volume change vs. gas transport) that are strongly related under normal conditions. Physiological factors such as variation in dead space, pendelluft, cardiac motion, and airways obstruction could cause the relationship between sVol and ventilation to quantitatively vary, as is seen with the differing slopes observed between animals and between different lung regions (Figs. 3 and 4). It is unknown what effect lung injury or disease would have on this correlation. In addition, registration errors and the scale-dependent nature of ventilatory heterogeneity will increase the noise of these measurements so that direct ROI-to-ROI comparisons will become increasingly noisier as the ROI size decreases. However, within these limitations, the technique may be useful for identifying patterns of ventilation and ventilation heterogeneity, particularly over large regions of the lung, and for estimating changes in regional volume change or volume strain with different patterns of ventilation.

In summary, there is a strong correlation of sVol to regional ventilation and many advantages to its use. In addition to the reduction in exposure to radiation and Xe side effects, images may be easily obtained from most conventional CT scanners. Regional ventilation for the entire lung may be estimated from whole lung CT image sets at two lung volumes. Ideally, gated images obtained during continuous ventilation, either prospectively gated to end-expiration and end-inspiration, as we obtained, or using new retrospective-gated reconstructions (13, 29, 31), will capture dynamic aspects of ventilation. These gating techniques are already being applied in human studies and will become more prevalent as scanner technology allows visualization of larger and larger volumes of the lung with each acquisition (20). However, even static whole lung images obtained from helical scans during breath-holds at two volumes will yield useful results (7) and provide a unique description of regional lung static mechanical properties that could be applied clinically to quantitate regional lung properties in patients already receiving this type of imaging, such as patients with severe emphysema being considered for lung volume reduction surgery or cystic fibrosis (2, 5, 17).

In conclusion, CT density changes over the respiratory cycle may be used to calculate the in situ sVol of identified lung regions. In normal lungs, regional sVol correlates well to regional specific ventilation measured with Xe-CT for ROIs from <1 to >4 ml volume and to regional volumes directly measured from implanted markers, albeit with increased noise as the region size falls. This technique may be useful for the noninvasive assessment of regional lung mechanical properties and prediction of ventilation distribution with significantly reduced exposure to radiation and Xe gas compared with the Xe-CT method. Regional sVol represents a safe and efficient surrogate for measuring regional ventilation in experimental studies and in patients.

	GRANTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
This study was supported by Department of Defense DAMD17-02-1-0732 and National Heart, Lung, and Blood Institute Grant HL-64368.

	ACKNOWLEDGMENTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 
We thank Jered Sieren and Earl Nixon for expert technical assistance and Drs. Junfeng Guo and Hidenori Shikata for the PASS and TSIA software.

	FOOTNOTES

 

Address for reprint requests and other correspondence: B. A. Simon, Dept. of Anesthesiology, Tower 711, The Johns Hopkins Hospital, Baltimore, MD 21287 (e-mail: bsimon{at}jhmi.edu)

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

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS ACKNOWLEDGMENTS REFERENCES

 

1. Abramoff MD, Magelhaes PJ, Ram SJ. Image processing with Image J. Biophotonics Intl 11: 36–42, 2004.
2. Becker MD, Berkmen YM, Austin JH, Mun IK, Romney BM, Rozenshtein A, Jellen PA, Yip CK, Thomashow B, Ginsburg ME. Lung volumes before and after lung volume reduction surgery: quantitative CT analysis. Am J Respir Crit Care Med 157: 1593–1599, 1998.[Web of Science][Medline]
3. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 327: 307–310, 1986.[CrossRef]
4. Chon D, Simon BA, Beck KC, Shikata H, Saba OI, Won C, Hoffman EA. Differences in regional wash-in and wash-out time constants for xenon-CT ventilation studies. Respir Physiol Neurobiol 148: 65–83, 2005.[CrossRef][Web of Science][Medline]
5. Dorlochter L, Nes H, Fluge G, Rosendahl K. High resolution CT in cystic fibrosis–the contribution of expiratory scans. Eur J Radiol 47: 193–198, 2003.[CrossRef][Web of Science][Medline]
6. Gattinoni L, Pelosi P, Crotti S, Valenza F. Effects of positive end-expiratory pressure on regional distribution of tidal volume and recruitment in adult respiratory distress syndrome. Am J Respir Crit Care Med 151: 1807–1814, 1995.[Abstract]
7. Guerrero T, Sanders K, Noyola-Martinez J, Castillo E, Zhang Y, Tapia R, Guerra R, Borghero Y, Komaki R. Quantification of regional ventilation from treatment planning CT. Int J Radiat Oncol Biol Phys 62: 630–634, 2005.[CrossRef][Web of Science][Medline]
8. Gur D, Drayer BP, Borovetz HS, Griffith BP, Hardesty RL, Wolfson SK. Dynamic computed tomography of the lung: regional ventilation measurements. J Comput Assist Tomogr 3: 749–753, 1979.[Web of Science][Medline]
9. Gur D, Shabason L, Borovetz HS, Herbert DL, Reece GJ, Kennedy WH, Serago C. Regional pulmonary ventilation measurements by xenon enhanced dynamic computed tomography: an update. J Comput Assist Tomogr 5: 678–683, 1981.[Web of Science][Medline]
10. Hamilton C, Stamey J. Using Bland-Altman to assess agreement between two medical devices: don't forget the confidence intervals! J Clin Monit Comp 21: 331–333, 2007.[CrossRef][Medline]
11. Hoffman EA, Reinhardt JM, Sonka M, Simon BA, Guo J, Saba O, Chon D, Samrah S, Shikata H, Tschirren J, Palagyi K, Beck KC, McLennan G. Characterization of the interstitial lung diseases via density-based and texture-based analysis of computed tomography images of lung structure and function. Acad Radiol 10: 1104–1118, 2003.[CrossRef][Web of Science][Medline]
12. Hubmayr RD, Walters BJ, Chevalier PA, Rodarte JR, Olson LE. Topographical distribution of regional lung volume in anesthetized dogs. J Appl Physiol 54: 1048–1056, 1983.[Abstract/Free Full Text]
13. Keall PJ, Starkschall G, Shukla H, Forster KM, Ortiz V, Stevens CW, Vedam SS, George R, Guerrero T, Mohan R. Acquiring 4D thoracic CT scans using a multislice helical method. Phys Med Biol 49: 2053, 2004.[CrossRef][Web of Science][Medline]
14. Kreck TC, Krueger MA, Altemeier WA, Sinclair SE, Robertson HT, Shade ED, Hildebrandt J, Lamm WJ, Frazer DA, Polissar NL, Hlastala MP. Determination of regional ventilation and perfusion in the lung using xenon and computed tomography. J Appl Physiol 91: 1741–1749, 2001.[Abstract/Free Full Text]
15. Lachmann B, Armbruster S, Schairer W, Landstra M, Trouwborst A, Van Daal GJ, Kusuma A, Erdmann W. Safety and efficacy of xenon in routine use as an inhalational anaesthetic. Lancet 335: 1413–1415, 1990.[CrossRef][Web of Science][Medline]
16. Li B, Christensen GE, Hoffman EA, McLennan G, Reinhardt JM. Establishing a normative atlas of the human lung: intersubject warping and registration of volumetric CT images. Acad Radiol 10: 255–265, 2003.[CrossRef][Web of Science][Medline]
17. Long FR, Williams RS, Adler BH, Castile RG. Comparison of quiet breathing and controlled ventilation in the high-resolution CT assessment of airway disease in infants with cystic fibrosis. Pediatr Radiol 35: 1075–1080, 2005.[CrossRef][Web of Science][Medline]
18. Marcucci C, Nyhan D, Simon BA. Distribution of pulmonary ventilation using Xe-enhanced computed tomography in prone and supine dogs. J Appl Physiol 90: 421–430, 2001.[Abstract/Free Full Text]
19. Martynowicz MA, Minor TA, Wilson TA, Walters BJ, Hubmayr RD. Effect of positive end-expiratory pressure on regional lung expansion of oleic acid-injured dogs. Chest 116: 28–29, 1999.[CrossRef]
20. Mori S, Endo M, Nishizawa K, Murase K, Fujiwara H, Tanada S. Comparison of patient doses in 256-slice CT and 16-slice CT scanners. Br J Radiol 79: 56–61, 2006.[Abstract/Free Full Text]
21. Pan Y, Kumar D, Hoffman EA, Christensen GE, McLennan G, Song JH, Ross A, Simon BA, Reinhardt JM. Estimation of regional lung expansion via 3D image registration. Proc SPIE 5746: 453–464, 2005.[CrossRef]
22. Robertson HT, Hlastala MP. Microsphere maps of regional blood flow and regional ventilation. J Appl Physiol 102: 1265–1272, 2007.[Abstract/Free Full Text]
23. Robertson HT, Kreck TC, Krueger MA. The spatial and temporal heterogeneity of regional ventilation: comparison of measurements by two high-resolution methods. Respir Physiol Neurobiol 148: 85–95, 2005.[CrossRef][Web of Science][Medline]
24. Simon BA. Non-invasive imaging of regional lung function using X-ray computed tomography. J Clin Monit Comput 16: 433–442, 2000.[CrossRef][Web of Science][Medline]
25. Simon BA. Regional ventilation and lung mechanics using X-ray CT. Acad Radiol 12: 1414–1422, 2005.[CrossRef][Web of Science][Medline]
26. Simon BA, Marcucci C, Fung M, Lele SR. Parameter estimation and confidence intervals for Xe-CT ventilation studies: a Monte Carlo approach. J Appl Physiol 84: 709–716, 1998.[Abstract/Free Full Text]
27. Tajik JK, Chon D, Won C, Tran BQ, Hoffman EA. Subsecond multisection CT of regional pulmonary ventilation. Acad Radiol 9: 130–146, 2002.[CrossRef][Web of Science][Medline]
28. Tajik JK, Tran BQ, Hoffman EA. Xenon enhanced CT imaging of local pulmonary ventilation. Proc SPIE 2709: 40–54, 1996.[CrossRef]
29. Vedam SS, Keall PJ, Kini VR, Mostafavi H, Shukla HP, Mohan R. Acquiring a four-dimensional computed tomography dataset using an external respiratory signal. Phys Med Biol 48: 45, 2003.[CrossRef][Web of Science][Medline]
30. Victorino JA, Borges JB, Okamoto VN, Matos GF, Tucci MR, Caramez MP, Tanaka H, Sipmann FS, Santos DC, Barbas CS, Carvalho CR, Amato MB. Imbalances in regional lung ventilation: a validation study on electrical impedance tomography. Am J Respir Crit Care Med 169: 791–800, 2004.[Abstract/Free Full Text]
31. Werner R, Handels H, Ehrhardt J, Saring D, Frenzel T, Lu W, Low D. Motion artifact reducing reconstruction of 4D CT image data for the analysis of respiratory dynamics. Methods Inf Med 46: 254–260, 2007.[Web of Science][Medline]
32. Yonas H, Grundy B, Gur D, Shabason L, Wolfson SK Jr, Cook EE. Side effects of xenon inhalation. J Comput Assist Tomogr 5: 591–592, 1981.[Web of Science][Medline]

This article has been cited by other articles:

				C. Salito, A. Aliverti, D. S. Gierada, G. Deslee, R. A. Pierce, P. T. Macklem, and J. C. Woods Quantification of Trapped Gas with CT and 3He MR Imaging in a Porcine Model of Isolated Airway Obstruction Radiology, November 1, 2009; 253(2): 380 - 389. [Abstract] [Full Text] [PDF]

				S. Bayat, S. Strengell, L. Porra, T. Z. Janosi, F. Petak, H. Suhonen, P. Suortti, Z. Hantos, A. R. A. Sovijarvi, and W. Habre Methacholine and Ovalbumin Challenges Assessed by Forced Oscillations and Synchrotron Lung Imaging Am. J. Respir. Crit. Care Med., August 15, 2009; 180(4): 296 - 303. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Free All Versions of this Article: 104/4/1177    most recent 00212.2007v1 Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (2) Citing Articles via Google Scholar Google Scholar Articles by Fuld, M. K. Articles by Simon, B. A. Search for Related Content PubMed PubMed Citation Articles by Fuld, M. K. Articles by Simon, B. A.