Quantitative analysis of computed tomography (CT) has been combined with a stereologically based histologic analysis of lung structure to assess regional lung inflation and the structural features of the lung parenchyma. In this study, CT measurements of lung inflation were compared with histologic estimates of surface area in order to develop prediction equations that allow lung surface to volume ratio and surface area to be predicted from an analysis of the CT scan. The results show that mild emphysema is associated with an increase in lung volume and a reduction in surface to volume ratio, whereas surface area and tissue weight were only decreased in severe disease. The CT predicted surface to volume ratio correlated with histology, and both predicted and measured surface areas correlated with the diffusing capacity. We conclude that this CT analysis can be used to monitor the progression of emphysematous lung destruction in individual patients, and to assess the impact of both surgical and medical treatments for emphysema.
Emphysema is defined as abnormal permanent enlargement of the airspaces distal to the terminal bronchioles, accompanied by destruction of their walls, without obvious fibrosis (1). These changes decrease the attenuation of X-rays passing through the thorax and shift their distribution, allowing advanced emphysema to be detected during life (2, 3).
The ability to estimate the extent and severity of emphysema during life is important for several reasons. Accurate detection of lung destruction when it first appears and careful mapping of its progression are required to understand the natural history of emphysema. The treatment of advanced disease by lung volume reduction surgery requires knowledge of the location of the lesions and an objective method of assessing the surgical result. Finally, the recent provocative experimental studies suggesting that alveolar number and surface to volume ratio can be restored by pharmocotherapy in rats with elastase-induced emphysema (4) suggest a future need for measurements that can accurately assess the effectiveness of such therapeutic interventions.
In this study the computed tomography (CT) scan and lung histology were used to assess the lungs of patients undergoing either lobar resection for tumor or lung volume reduction surgery. We hypothesized that CT data could be used to measure the weight, tissue, and gas volume of the lung, and to predict lung surface to volume ratios and surface area in living patients. These CT-based predictions may be useful in assessing surgical and medical interventions as well as in following the natural history of emphysema.
This study was approved by the ethical review boards of St. Paul's Hospital, the University of Pittsburgh Medical Center, and the University of British Columbia. Informed consent was obtained for the use of physiologic data, CT scans, and the surgically resected tissue. The patients in the control and mild emphysema groups required either a lobectomy or pneumonectomy for a nonobstructing, peripheral bronchogenic carcinoma, less than 3 cm in diameter, and were part of an ongoing study of lung structure and function at the University of British Columbia. The severe emphysema group was selected for lung volume reduction surgery at the University of Pittsburgh using previously defined criteria (5). The separation into three groups was based on the percent of the lung that was determined to be emphysematous using the “density mask” technique (3).
Spirometry and lung volumes were measured preoperatively with the subjects seated in a volume displacement body plethysmograph using previously described techniques (5-8). FRC was measured using the Boyle's Law technique (9-11). Total lung capacity (TLC) was calculated by adding inspiratory capacity (IC) to FRC, and residual volume (RV) was calculated by subtracting vital capacity (VC) from TLC. Diffusing capacity (Dl CO) was measured by the single-breath method of Miller and associates (12). The predicted normal values for FEV1, FVC, and Dl CO were those of Crapo (13), and TLC was predicted using Goldman and Becklake's values (14).
All subjects received a conventional, noncontrast CT scan (10-mm thick contiguous slices) on a GE 9800 Highlight Advantage CT scanner (General Electric Medical Systems, Milwaukee, WI) approximately 1 wk prior to surgery. The scanners were calibrated regularly using standard water and air phantoms to allow for comparisons between individuals. These scans were performed with the subject supine while breath-holding.
The CT scan analysis used to evaluate the lung has been described in detail elsewhere (6, 7). Briefly, it was performed using a program written for the numerical analysis package PV-Wave (Visual Numerics, Boulder, CO). The lung parenchyma was segmented from the chest and the large central blood vessels using CT values of −1,000 to −500 Hounsfield units (HU), and the volume of the whole lung (tissue and airspace) was calculated by summing the voxel dimensions in each slice. The density of the lung (g/ml) was estimated by adding 1,000 to the Hounsfield units of each voxel, and dividing by 1,000 (6, 15). Lung weight was estimated by multiplying the mean lung density by the volume. The volume of gas per gram of tissue for each voxel was calculated according to the equation (6, 7, 16):
where specific volume is the inverse of density. The density of the lung (tissue and gas) was measured from the CT, and the density of tissue was assumed to be 1.065 g/ml (16). The frequency distribution of the milliliter of gas per gram of tissue for the individual voxels was plotted and the moments of the curve were obtained.
A CT estimation of tissue volume fraction was calculated according to Equation 2:
and used to correct the histologic estimates of the tissue and airspace to the level of inflation the patient achieved during the CT scan (6).
The inflation of each lung voxel was categorized as to whether it was below previously reported values of TLC (< 6.0 ml/g) (7), between TLC and a density mask value chosen by Miller and colleagues to define emphysema (6.0–10.2 ml/g) (3), and inflation values larger than this cut-off (> 10.2 ml/g), which identifies lesions greater than 5 mm in diameter. Control patients were defined as having less than 5% of their lung volume above 10.2 ml/g, patients with mild disease were those with greater than 5% but less than 20% of their lung volume above 10.2 ml/g, while patients who had severe emphysema were defined as those in whom greater than 20% of their lung was above this volume. The accuracy of this technique for measuring emphysema was determined by comparing the CT measurement in a lobe that was to be resected with the morphometrically determined amount of emphysema in the resected lobe.
The resected specimens were obtained directly from the operating room and inflated with Optimal Cutting Temperature compound (Miles Laboratories, Elkhart, IN) that was diluted 1:1 with normal saline and frozen over liquid nitrogen. The frozen specimen was cut into 2-cm thick slices in the transverse plane on a band saw. Cores were removed from the lung and stored at −70° C for other purposes, and the remainder of the lung tissue was transferred to 10% buffered formalin and fixed at room temperature for at least 24 h (6, 7). The tissue from the patients undergoing lung volume reduction surgery for severe emphysema was received directly from the operating room and fixed, without inflation, in 10% formalin. Hematoxylin- and eosin-stained histologic sections were prepared from random samples of the surgical specimens.
To optimize the sampling for the stereologic analysis, a cascade design technique was used as has been previously described (6, 7, 18). Level 1 was performed on the fixed slices of the resected specimens by floating them in water and overlaying a grid of points (19). The slices were examined with the aid of a 2× magnifying lens (19) and the number of points falling on emphysematous lesions (larger than 5 mm diameter, smaller than 5 mm diameter), and normal lung parenchyma were counted. Levels 2 and 3 were performed on all available sections at the light microscopy level using the point counting program Gridder (Wilrich Tech, Vancouver, BC), which generated random fields of view, projected a grid on to the field of view via a camera-lucida attachment on a Nikon Labophot light microscope and tabulated the counts. Level 2 was done at ×100 magnification using a Weibel multipurpose test grid with 40 lines and 80 points (20). The number of points falling on airspace, tissue (lung parenchyma), and medium-sized blood vessels (50–1,000 μm) as well as the number of intersects between the grid lines and the parenchymal-airspace interface were tabulated. Level 3 was performed on 10 random fields of view per slide at ×400 magnification and the number of points falling on airspace components (alveolar macrophages, alveolar polymorphonuclear leukocytes, alveolar fluid, and empty space) as well as tissue components (alveolar wall, capillary lumen, and small blood vessels [20–50 μm]) were counted using a 100-point grid.
The volume fraction (VV) of each of the lung components, for example, tissue (Vv (tis)), were estimated at each level according to Equation 3:
where ΣP(tis) is the number of points falling on tissue, and ΣP(total) is the total number of points counted, tissue and non-tissue.
The overall VV is calculated by multiplying the VV of the lung component at the highest level by the VV in the previous levels; for example, the volume fraction of capillaries would be calculated as follows:
The surface density of the tissue–air interface, Sv (tis), was calculated using the following equation:
where l is the length of the grid unit line, ΣI is the number of intersects counted, and ΣP is the number of line end points falling on parenchymal tissue. Since surface density is the surface area in a given volume, the surface area of the tissue is calculated by multiplying the surface density by the volume fraction of tissue (obtained above) by the lung volume.
The median value of the morphologically determined surface area per volume was compared with the median CT measurements of ml gas per gram of tissue and the regression equation of best fit was determined. This equation was then used to predict the surface to volume ratio and surface area for each voxel, and these values were summed to obtain the value for the entire lung. The histologically measured and CT predicted surface areas were then correlated with each other and independently with the diffusing capacity (Dl CO).
The data were analyzed using one-way analysis of variance and multivariate analysis of variance. Transformations were made on certain variables to normalize distributions and to make variances homogeneous. A p value of less than 0.05 was considered significant.
The patient demographics (Table 1) show that there are more males than females in all three groups, and there is no difference in body size (height and weight), but the control group is slightly younger than the other two groups. The smoking history is not different between the three groups (p > 0.05), but the patients who have mild emphysema tended to smoke less. Those with mild emphysema have FEV1, FVC, Dl CO, TLC, and RV values that are similar to the control group, but the FEV1/FVC ratio is reduced and the FRC is elevated. Those with severe emphysema have grossly abnormal lung function characterized by reduced FEV1, FVC, FEV1/FVC, and Dl CO and elevated TLC, FRC, and RV.
|Control||Mild Emphysema||Severe Emphysema|
|Age, yr||59 ± 2||66 ± 3*||66 ± 2*|
|Smoking, pack yr||45 ± 6||28 ± 6||58 ± 9|
|Weight, kg||70 ± 2||77 ± 4||69 ± 3|
|Height, cm||167 ± 2||175 ± 3||168 ± 3|
|FEV1, % predicted||93 ± 3||84 ± 5||28 ± 2†|
|FVC, % predicted||102 ± 3||102 ± 6||72 ± 4†|
|FEV1/FVC||0.73 ± 0.01||0.64 ± 0.04*||0.30 ± 0.01†|
|Dl CO, % predicted||89 ± 3||78 ± 13||31 ± 4†|
|TLC, % predicted||113 ± 3||122 ± 6||129 ± 5*|
|FRC, % predicted||128 ± 5||157 ± 9*||188 ± 9†|
|RV, % predicted||133 ± 8||154 ± 10||222 ± 13†|
The CT estimates of lung volume (Table 2) show a higher total lung and airspace volume in the severe compared with the mild emphysema group, which in turn is greater than the control group. The tissue volume and therefore the lung weight is less in the group with severe emphysema than in the control group, but there is no difference in lung weight between the mild emphysema group and the control subjects. The quantitative histology shows that there is a significant decrease in the surface to volume ratio of lung in mild emphysema and a very marked reduction in this value in severe emphysema.
|Control||Mild Emphysema||Severe Emphysema|
|Total lung volume, ml||4,772 ± 223||6,232 ± 410*||6,725 ± 384†|
|Airspace volume, ml||3,815 ± 194||5,195 ± 388*||5,964 ± 353*|
|Tissue volume, ml||957 ± 34||1,037 ± 33||760 ± 35†|
|Lung weight, g||1,019 ± 37||1,104 ± 35||810 ± 37†|
|% Voxels < 6.0 ml/g||79 ± 3||52 ± 4*||25 ± 1†|
|% Voxels 6.0–10.2 ml/g||20 ± 3||35 ± 4*||28 ± 1*|
|% Voxels > 10.2 ml/g||1 ± 0||13 ± 1*||47 ± 2†|
|Measured surface area/volume, cm2/ml||256 ± 24||165 ± 23*||43 ± 6†|
|Predicted surface area/volume, cm2/ml||300 ± 9||212 ± 12*||138 ± 7†|
|Measured surface area, m2||118 ± 11||97 ± 8||30 ± 5†|
|Predicted surface area, m2||128 ± 5||119 ± 3||60 ± 3†|
The frequency distribution curves (Figure 1) for the ml gas per gram of tissue present in each CT voxel are different between the three groups. The values for the normal TLC (6.0 ml/g) and the cut-off at −910 HU in the density mask technique (10.2 ml/g) are indicated by vertical arrows. In the control group 79 ± 3% of the voxels are below 6.0 ml/g, 20 ± 3% are between 6.0 ml/g and 10.2 ml/g and 1 ± 0 are above 10.2 ml/g. The mild cases have 52 ± 4, 35 ± 4, and 13 ± 1% of their lung voxels in these three categories, while the severe cases showed 25 ± 1, 28 ± 1, and 47 ± 2%, respectively (Table 2). In those with severe emphysema, 20% of the lung CT voxels are greater than 20 ml/g, which is more than three times the amount of air contained in the normal human lung at TLC (7).
|Control (n = 23)||Mild Emphysema (n = 7)|
|0.0–6.0 ml/g||Normal lung||80 ± 8||75 ± 8||53 ± 12||49 ± 10|
|6.0–10.2 ml/g||Lesions < 5 mm||20 ± 8||25 ± 8||30 ± 9||37 ± 3|
|> 10.2 ml/g||Lesions > 5 mm||1 ± 1||0 ± 0||18 ± 6||14 ± 6|
Additional data derived from this frequency distribution show that the control lungs have a symmetrical distribution with mean, median, and mode values that are closely similar (4.5 ± 0.2, 4.4 ± 0.1, and 4.4 ± 0.2 ml/g) and a relatively small variance (2.9 ± 2.3 ml/g). The mild emphysema group shows a distribution that is slightly shifted and skewed to the right around a mean of 7.1 ± 0.3 ml/g, median of 5.8 ± 0.3 ml/g, and mode of 5.3 ± 0.6 ml/g with a very large variance (208 ± 112 ml/g). The severe emphysema group has a flattened distribution that is markedly shifted and skewed to the right with a mean of 14.0 ± 1.2 ml/g, a median of 9.8 ± 0.4 ml/g, mode of 8.1 ± 0.5 ml/g, and the largest variance (579 ± 200 ml/g). When the median values are expressed as percentage of measured ml/g at TLC, the control and mild emphysema patients are not different (66.0 ± 2.2% versus 74.6 ± 3.2%); however, the median value for the patients who had severe emphysema is significantly higher (99.7 ± 4.9%).
A comparison of the amount of emphysema detected in the same lobe by both CT and point counting of the resected specimen (Table 3) shows that the volume fraction of lesions greater than 5 mm in diameter measured by morphometry is similar to the fraction of lung inflated above 10.2 ml/g. It also shows that lesions less than 5 mm in diameter correspond to the fraction of the lung inflated between 6.0 and 10.2 ml/g and that regions inflated below 6.0 ml/g are normal in appearance.
Figure 2A shows the relationship between volume of gas per gram of tissue measured using the CT and the surface to volume ratio measured from the histologic material. Each data point represents the median value for both variables in an individual case. The equation:
which best describes these data, was used to predict surface area to volume ratio for each voxel in the lung. The surface area was estimated by multiplying the volume of the voxel by the predicted surface to volume ratio. The total surface area of the lung can then be estimated by summing the values for each voxel. A comparison of the estimate of surface area to volume obtained using the CT data and the prediction equation to that determined histologically (Figure 2B) shows a good correlation (r2 = 0.75; p < 0.05) between the two values. However, the regression line is different from the line of identity, indicating a tendency for the CT to overestimate the surface area to volume in the cases with lower histologically determined values.
Figure 3 shows the regression analysis of both the histologically estimated and CT predicted surface area versus measured Dl CO. The regression using the histologically measured values has a positive slope of 0.1 ml/min/mm Hg/m2, an intercept of 7.1 ml/min/mm Hg and an r2 of 0.59. The r2 for the CT estimate of surface area is 0.57 with a slope of 0.2 ml/min/ mmHg/m2 and an intercept of 0.41 ml/min/mm Hg.
The results show that CT data can be used to calculate total lung weight and the tissue and airspace volume of the lung. It also shows that the prediction equation arising from the comparison of the CT measurements of lung inflation (ml/g) and morphologic assessment of surface to volume ratio can be used to estimate total and regional surface to volume ratios and lung surface area from the CT scan. It has been shown that Dl CO correlates best with morphologic measurements of alveolar surface area per volume (2, 21), despite the fact that Dl CO alone is nonspecific for the diagnosis of emphysema (22). Our data show that both the histologically estimated, which can be considered to be the gold standard, and the CT predicted lung surface areas correlate with the measured Dl CO (Figure 3). This suggests that the relatively noninvasive CT estimate of surface area provides a tool for tracking the destruction of lung surface area by emphysema. The major advantage of this new technique is that in addition to providing data concerning overall lung destruction it also identifies the specific locations in the lung where the surface has been destroyed.
The simple calculation used to convert the HU of each voxel to the more physiologically meaningful unit, milliliters of gas per gram of tissue, allows easy calculation of regional and total lung volume and a rapid determination of the overall lung volume at which the CT scan was obtained (7). Further analysis showed that the lung from control subjects had a symmetrically distributed regional lung inflation with a similar mean, median, and mode where 99% of the lung voxels fell below a density mask cut-off value of 10.2 ml/g (−910 HU ) (Table 3, Figure 1). The presence of emphysema changed the shape of this distribution by shifting the curve to the right as well as skewing it so that the more severe the extent of disease, the higher the proportion of the lung found above 10.2 ml/g (Table 3, Figure 1). This cut-off value is three standard deviations above that obtained by dividing measured TLC by measured lung weight in control lungs (6.0 ± 1.1 ml/g) (7). When the lung volume at which the CT scan was obtained is expressed as a percent of measured TLC, there is a progression toward TLC as disease severity increased (control, 66.0 ± 2.2%; mild, 74.6 ± 3.2; and severe, 99.7 ± 4.9%), which is consistent with the pressure–volume characteristics of emphysematous lesions (23).
The comparison of CT to morphology in the resected lobes confirms previous reports (3, 17) that the standard density mask cut-off (−910 HU or 10.2 ml/g) fails to detect emphysematous lesions less than 5 mm in diameter (Table 3). This comparison also shows that these smaller lesions are located in regions of the lung with values between normal TLC (6.0 ml/g) and the density mask cut-off (10.2 ml/g). This shows that the present technique provides an advantage over those that rely on a single cut-off value (2, 3, 17, 24) because it identifies lung tissue containing the smaller lesions.
Figure 2A shows that although there is intersubject variability, there is an exponential relationship between the median values of volume of gas per gram of tissue obtained from the CT scan and the surface to volume ratio obtained using histology. Sampling of the lung for quantitative histology requires that either the lung structure is uniform throughout or that an appropriate sample method is used (25). We have made an assumption that the histologic specimens are representative of the whole lung structure because we were limited in the availability of tissue specimens. This assumption is reasonable in the heterogeneous distribution of emphysematous lesions in the mild cases because both CT and histology sampled an entire lobe. However, in the severe cases where only the resected regions of the lung are sampled, a bias is introduced. This is evident in Figure 2B, where the correlation between histology estimates and CT predicted surface area per volume is different from the line of identity, and all of the data points from patients with severe emphysema are above the line. However, the correlation between histology and CT estimates of volume of gas per gram of tissue (r2 = 0.75) suggests that this prediction is reasonable. Data from a greater number of patients with varying degrees of emphysema should increase the predictive ability of this equation.
Mild emphysema was associated with no change in surface area but an increase in lung volume (Table 2), which is consistent with a reduction in elastic recoil. The Laplace relationship between pressure and radius shows that as alveolar diameter increases, the recoil pressure decreases (26), and since the surface area of the lung is conserved in mild disease, the observed decrease in surface to volume ratio must be due to an enlargement of the alveoli. As the surface area of the lung is conserved, the hyperinflation could be due to proteolytic destruction of elastic fibers within the alveolar walls, which is directly associated with alveolar enlargement (26). Alternatively, the lung overinflation could be due to prolongation of the time constants of the peripheral lung units due to disease obstructing the peripheral airways (27). Unfortunately, it is not possible to distinguish between these possibilities from the present data.
We have previously shown that CT scans can be used to estimate the volume fraction of tissue and airspace in a group of control patients and patients with idiopathic pulmonary fibrosis (6, 7). The present study extends these observations to show that emphysematous lung destruction can be quantified by analyzing the electronic record of a relatively noninvasive CT scan and using the data in a prediction equation that relates CT estimates of lung inflation to histologic measurements of surface to volume ratio. The algorithms that have been developed make estimates of surface area to volume ratios and surface area available to clinicians to assess both the natural history of emphysema and the impact of surgical (28) and medical interventions (29).
The writers thank Hayedeh Behzad and Bernard Meshi for technical assistance, Barbara Moore for collecting the pulmonary function data from the resected patients, the CT/MRI staff at St. Paul's Hospital and Brad Rogers at the University of Pittsburgh for their patience in gathering and transferring the patients' CT images, and Dr. Samuel Yousem at the University of Pittsburgh for preparation of the lung volume reduction surgery specimens.
Supported by Medical Research Council of Canada Grant 4219, the National Centres of Excellence in Respiratory Health, and the George H. Love Foundation.
|1.||Snider G. L., Kleinerman J., Thurlbeck W. M., Bengali Z. H.The definition of emphysema. Report of the National Heart, Lung and Blood Institute, Division of Lung Diseases Workshop. Am. Rev. Respir. Dis.1321985182185|
|2.||Gould G. A., MacNee W., McLean A., Warren P. M., Redpath A., Best J. J., Lamb D.CT measurements of lung density in life can quantitate distal airspace enlargement: an essential defining feature of human emphysema. Am. Rev. Respir. Dis.1371988380392|
|3.||Müller N. L., Staples C. A., Miller R. R., Abboud R. T.‘Density mask': an objective method to quantitate emphysema using computed tomography. Chest941988782787|
|4.||Massaro G. D., Massaro D.Retinoic acid treatment abrogates elastase-induced pulmonary emphysema in rats. Nature Med.31997675677|
|5.||Sciurba F. C., Rogers R. M., Keenan R. J., Slivka W. A., Gorcsan J., Ferson P. F., Holbert J. M., Brown M. L., Landreneau R. J.Improvement in pulmonary function and elastic recoil after lung- reduction surgery for diffuse emphysema. N. Engl. J. Med.334199610951099|
|6.||Coxson H. O., Hogg J. C., Mayo J. R., Behzad H., Whittall K. P., Schwartz D. A., Hartley P. G., Galvin J. R., Wilson J. S., Hunninghake G. W.Quantification of idiopathic pulmonary fibrosis using computed tomography and histology. Am. J. Respir. Crit. Care Med.155199716491656|
|7.||Coxson H. O., Mayo J. R., Behzad H., Moore B. J., Verburgt L. M., Staples C. A., Paré P. D., Hogg J. C.The measurement of lung expansion with computed tomography and comparison with quantitative histology. J. Appl. Physiol.79199515251530|
|8.||Hogg J. C., Wright J. L., Wiggs B. R., Coxson H. O., Saez A. O., Paré P. D.Lung structure and function in cigarette smokers. Thorax491994473478|
|9.||DuBois A. B., Botelho S. Y., Bedell G. W., Marshall R., Comroe J. H.A rapid plethysmographic method for measuring thoracic gas volume: a comparison with nitrogen washout method for measuring functional residual capacity in normal subjects. J. Clin. Invest.351956322326|
|10.||McCuaig K. E., Vessal S., Coppin C., Wiggs B. R., Dahlby R., Paré P. D.Variability in measurements of pressure–volume curves in normal subjects. Am. Rev. Respir. Dis.1311985656658|
|11.||Osborne S., Hogg J. C., Wright J. L., Coppin C., Paré P. D.Exponential analysis of the pressure-volume curve: correlation with mean linear intercept and emphysema in human lungs. Am. Rev. Respir. Dis.137198810831088|
|12.||Miller A., Thornton J. C., Warshaw R., Anderson H., Teirstein A. S., Selikoff I. J.Single breath diffusing capacity in a representative sample of the population of Michigan, a large industrial state: predicted values, lower limits of normal, and frequencies of abnormality by smoking history. Am. Rev. Respir. Dis.1271983270277|
|13.||Crapo R.Spirometric values using techniques and equipment that meet ATS recommendations. Am. Rev. Respir. Dis.1231981659664|
|14.||Goldman H. I., Becklake M. R.Respiratory function test: normal values at median altitudes and the prediction of normal results. Am. Rev. Tuberc.791959457467|
|15.||Hedlund L. W., Vock P., Effmann E. L.Evaluating lung density by computed tomography. Semin. Respir. Med.519837687|
|16.||Hogg J. C., Nepszy S.Regional lung volume and pleural pressure gradient estimated from lung density in dogs. J. Appl. Physiol.271969198203|
|17.||Miller R. R., Müller N. L., Vedal S., Morrison N. J., Staples C. A.Limitations of computed tomography in the assessment of emphysema. Am. Rev. Respir. Dis.1391989980983|
|18.||Cruz-Orive L. M., Weibel E. R.Sampling designs for stereology. J. Microsc.1221981235257|
|19.||Thurlbeck W. M.Internal surface area and other measurements in emphysema. Thorax221967483496|
|20.||Aherene, W. A., and M. S. Dunnill. 1982. Morphometry. Edward Arnold, London. 46–58.|
|21.||McLean A., Warren P. M., Gillooly M., MacNee W., Lamb D.Microscopic and macroscopic measurements of emphysema: relation to carbon monoxide gas transfer. Thorax471992144149|
|22.||MacNee W., Gould G., Lamb D.Quantifying emphysema by CT scanning: clinicopathologic correlates. Ann. N.Y. Acad. Sci.6241991179194|
|23.||Hogg J. C., Nepszy S. J., Macklem P. T., Thurlbeck W. M.Elastic properties of the centrilobular emphysematous space. J. Clin. Invest.48196913061312|
|24.||Gevenois P. A., de Maertelaer V., De Vuyst P., Zanen J., Yernault J. C.Comparison of computed density and macroscopic morphometry in pulmonary emphysema. Am. J. Respir. Crit. Care Med.1521995653657|
|25.||Bolender R. P., Hyde D. M., Dehoff R. T.Lung morphometry: a new generation of tools and experiments for organ, tissue, cell, and molecular biology. Am. J. Physiol.2651993L521L548|
|26.||Haber, P. S., H. J. H. Colebatch, C. K. Y. Ng, and I. A. Greaves. 1983. Alveolar size as a determinant of pulmonary distensibility in mammalian lungs. J. Appl. Physiol.: Respir. Environ. Exerc. Physiol. 54:837–845.|
|27.||Hogg J. C., Macklem P. T., Thurlbeck W. M.Site and nature of airway obstruction in chronic obstructive lung disease. N. Engl. J. Med.278196813551360|
|28.||Rogers R. M., Sciurba F. C., Keenan R. J.Lung reduction surgery in chronic obstructive lung disease. Med. Clin. North Am.801996623644|
|29.||Crystal R. G.α1-Antitrypsin deficiency, emphysema, and liver disease: genetic basis and strategies for therapy. J. Clin. Invest.85199013431352|