Chronic obstructive pulmonary disease (COPD) is characterized by the presence of airflow obstruction caused by emphysema or airway narrowing, or both. Low attenuation areas (LAA) on computed tomography (CT) have been shown to represent macroscopic or microscopic emphysema, or both. However CT has not been used to quantify the airway abnormalities in smokers with or without airflow obstruction. In this study, we used CT to evaluate both emphysema and airway wall thickening in 114 smokers. The CT measurements revealed that a decreased FEV1 (%predicted) is associated with an increase of airway wall area and an increase of emphysema. Although both airway wall thickening and emphysema (LAA) correlated with measurements of lung function, stepwise multiple regression analysis showed that the combination of airway and emphysema measurements improved the estimate of pulmonary function test abnormalities. We conclude that both CT measurements of airway dimensions and emphysema are useful and complementary in the evaluation of the lung of smokers.
Chronic obstructive pulmonary disease (COPD) is characterized by the presence of airflow obstruction caused by emphysema or airway narrowing, or both (1). Cigarette smoking, which is the major risk factor of COPD, causes the abnormalities by inducing inflammation in the lung parenchyma and peripheral airways.
A large number of studies have been done to detect and quantify pulmonary emphysema using computed tomography (CT) (2-18). Low attenuation areas (LAA) on CT scans in vivo have been shown to represent macroscopic or microscopic emphysematous changes in the lungs of patients (2-7). Although the CT measurement of LAA correlates well with diffusing capacity, the relationship to measurements of airflow obstruction is less significant (5-15) presumably because airflow obstruction is related to both loss of recoil and inflammatory narrowing of the airways.
Recently some studies have been done to measure airway dimensions using CT (19-27). The measurement of airway wall thickness by high-resolution CT (HRCT) in patients with asthma has been shown to correlate with the severity of asthma (22). It has been suggested that CT has potential power to evaluate airways in patients with obstructive pulmonary disease (28). However, to our knowledge, there are no published CT studies of airway dimensions and comparison with pulmonary function tests (PFTs) in smokers with or without airflow obstruction.
We hypothesized that both the measurement of emphysema and airway dimensions using CT in smokers would correlate with PFTs. We also hypothesized that the measurements of airway dimensions by CT would provide additive value to the measurement of emphysema in the prediction of lung function abnormalities. In the present study, we used HRCT to quantify emphysema and thin-section helical (spiral) CT to quantify airway dimensions for comparison with measurements of PFTs in 114 smokers.
A total of 114 smokers (94 patients with COPD from Kyoto University Hospital and 20 asymptomatic volunteers) were recruited. A diagnosis of COPD was made according to the criteria of the American Thoracic Society (1). No patients with large bullae or other lung diseases were included in the study. All the patients were clinically stable at the time of the study, and inhalation of β-receptor agonist or anticholinergic drugs, or both, were withheld for at least 12 h before the study. None of the asymptomatic volunteers had a history of respiratory illness and none had respiratory symptoms at the time of the study. All subjects had PFTs and CT scans on the same day. This study was approved by the ethics committee of Kyoto University and informed consent was obtained from each participant.
PFTs were performed with a Chestac-65V (Chest MI Corp., Tokyo, Japan). Forced vital capacity (FVC), forced expiratory volume in one second (FEV1), residual volume (RV) using the helium dilution method, peak expiratory flow rate (PEFR), total lung capacity (TLC), carbon monoxide diffusing capacity (Dl CO), and alveolar volume (Va) using the single-breath method were measured. FVC, FEV1 and PEFR were expressed as percentage of predicted values (FVC%P, FEV1%P, and PEFR%P, respectively) according to the prediction equations of the Japanese Society of Chest Diseases (29). The ratio of FEV1 to FVC (FEV1/FVC) and the ratio of RV to TLC (RV/TLC) were expressed as percentages. The ratio of Dl CO to Va (Dl CO/Va) was expressed as the actual measurement values.
We used HRCT to quantify LAA and thin-section helical (spiral) CT to quantify airway dimensions. Both LAA and airway analyses could be done on a helical scan but because there are no available previous data for LAA using helical CT, we used both HR and helical CT in the present study. Both scans were performed in a supine position using the same CT scanner (X-Vigor; Toshiba, Tokyo, Japan). No contrast media were used. CT scans were obtained after deep inspiration. This minimized the influence of variable hyperinflation, allowed us to compare the results under the same conditions, and optimized breath-holding time.
The HRCT was performed using 2-mm collimation, scan time 1.0 s, 120 kVp, and 200 mA. The images were reconstructed on a 32-cm field of view (FOV) using a lung algorithm (FC83). Three HRCT scans were used for determination of LAA; a cranial (upper) section was obtained 1 cm above the superior margin of the aortic arch, a middle section was taken at 1 cm below the carina, and a caudal (lower) section was taken approximately 3 cm above the top of the diaphragm (8, 14-17).
The helical scan was performed using 120 kVp, 50 mA, 3-mm collimation, and pitch 1.0. Images were reconstructed using the FC10 algorithm at 2-mm spacings. A targeted reconstruction of the right lung was performed using a subject-specific FOV (153 to 214 mm). We used thin-section helical CT for the analysis of airways to accurately select the same anatomic location (origin of the apical bronchus of the right upper lobe) and to reduce the subject's radiation dose.
Each HRCT or helical CT image was composed of a 512 × 512 matrix of numeric data (CT numbers) in Hounsfield units (HU). These CT data were transferred to a PowerPC personal computer via a magneto-optical disk, and were analyzed using custom software written in C programming language (Symantec C++, Symantec Corp., CA).
The three HRCT scans were used for the analysis of LAA. Using a method previously reported (8, 14-17), the percentage of low attenuation pixels (LAA%) was calculated automatically. The cutoff level between normal lung density and LAA was defined as −960 HU (8, 14-18).
It has been reported that it is useful to divide the patients with COPD into two groups; one with LAA% less than 30 and the other with LAA% more than 30 (15). Thus, we divided the subjects into three groups; COPD patients with LAA% less than 30%, COPD patients with LAA% more than 30%, and the asymptomatic smokers. We designated these three groups as COPD ⩽ 30, COPD > 30, and Asymptomatics, respectively.
The trunk of the apical bronchus of the right upper lobe is usually sliced in cross section and easily identified on CT scans. This airway was chosen because it was large enough to be accurately measured and anatomically reproducible between subjects (see Validation of the Airway Analysis). Images containing the apical bronchus to the right upper lobe were selected by a consensus reading of three pulmonologists (Y.N., S.M., and H.S.). With the bronchus identified, the following parameters were measured automatically on the computer (Figure 1); luminal area (Ai), short radius (SR), and long radius (LR) of the lumen, and airway wall thickness (T). Details are described in the . In brief, the following procedures were performed: (1) The lumen of the bronchus was identified using a threshold −500 (27). The area of the lumen was considered as Ai. (2) SR and LR were defined as the shortest and longest distances from the centroid point of the lumen to the edge. (3) From the centroid point of the lumen, 128 rays fanning out over 360° were examined to determine T along the rays using the “full-width at half-maximum” principle (25, 26). (4) Those rays which projected onto the adjacent vessel were excluded. T was calculated from the nonexcluded rays.

Fig. 1. A representative case showing the process of airway analysis. Using a helical CT image containing the apical bronchus to the upper lobe (A), the algorithm defines the luminal area (B) and wall thickness (C). Note that the identification of the airway wall thickness was successfully performed even when the pulmonary blood vessel ran parallel to the bronchus.
[More] [Minimize]Because each subject had one to five images (2.6 ± 1.0, mean ± SD) through the apical sectional bronchus, airway and wall thickness measurements were averaged to yield final Ai and T. Assuming that, in a cross-sectional plane, the airway lumen is a true circle and T is constant throughout the wall, the total diameter (D) of the bronchus was calculated as D = 2 + 2T. We calculated the outer area of the bronchus (Ao) from D; Ao = š (D/2)2. Airway wall area (WA) was calculated as Ao − Ai. We also used two previously reported indices to compare airway dimensions: the ratio of airway wall thickness to total diameter (the T/D ratio) (20-22) and the percentage wall area [WA% = (WA/Ao) × 100] (19, 22).
To test the validity of using the airway dimensions of a single airway as a measure of airway thickening, we chose 20 subjects randomly from the total of 114 smokers and measured airway dimensions on all the airways that were sectioned in cross-section in the helical CT scans. The average WA%, for the airways other than the apical segmental bronchus, was calculated for each individual and compared with the WA% of the apical segmental bronchus.
The airway analysis software and algorithm were tested using a phantom. The phantom was made from a polystyrene foam block and eight plastic tubes, which represent lung parenchyma and airways respectively. The actual tube dimensions were measured using an optical micrometer caliper to the closest 0.01 mm. The plastic tubes ranged from 0.55 mm to 2.25 mm in wall thickness and from 8.0 mm2 to 335.1 mm2 in luminal area. Helical scans of the phantom were obtained and reconstructed on a 200-mm FOV. The pixel size at this FOV is 0.39 × 0.39 mm.
All statistical analyses were done using Stat View software (SAS Institute Inc., Cary, NC). Results were expressed as mean ± SD. Univariate (linear) regression analysis and stepwise multiple regression analysis were used to evaluate the relationship between the CT parameters and PFTs. Mann-Whitney U test and analysis of variance (ANOVA) with Bonferroni/Dunn correction were used to compare groups. Adjustments for multiple comparisons (Bonferroni adjustments) were performed in both the univariate and the multivariate correlations. A value of p less than 0.05 was considered significant.
The maximal CT density of the plastic tubes was 49 ± 102 HU, which was virtually identical with the 50 ± 49 HU in the 10 randomly selected images of the right apical bronchus. There were significant positive relationships between the actual and CT-measured wall thickness (r = 0.975, p < 0.0001) and between the actual and CT-measured luminal area (r = 1.00, p < 0.0001). Figure 2 shows the error of CT measurements expressed as follows:

Fig. 2. Percent error for luminal area (A) and wall thickness (B) in the phantom study. The error of the smallest plastic tube (0.55 mm in thickness and 8.0 mm2 in actual area) was 27% in area and −81% in thickness. However, the errors of other phantom tubes were within the range of −2% to 4% in area and −11% to 9% in thickness.
[More] [Minimize]Error (%) = (actual measurement − CT measurement)/ actual measurement × 100
The results showed that our system and algorithm are accurate within the range of 1.0 mm to 2.3 mm in thickness and within the range of 12 mm2 to 335 mm2 in luminal area.
All the subjects were male and were either current (53/114) or ex-smokers. Anthropometric and pulmonary function data are shown in Table 1. The results of CT measurements are shown in Table 2. Because body size varied considerably, T, Ai, and Ao were normalized using body surface area (BSA) and . BSA was chosen because of the similarity of the units (mm2 for airway walls and m2 for BSA). However, similar results were obtained when we normalized airway measurements with height. The mean SR/LR ratio (airway roundness ratio) was 0.86 ± 0.11.
Mean ± SD | Range | |||
---|---|---|---|---|
Age, yr | 68 ± 9 | 32–88 | ||
Smoking index, pack-years | 71 ± 43 | 5–268 | ||
Height, cm | 162 ± 7 | 145–180 | ||
Weight, kg | 57 ± 8 | 40–80 | ||
Body surface area, m2 | 1.61 ± 0.13 | 1.29–1.92 | ||
FVC, %predicted | 74 ± 20 | 35–127 | ||
FEV1, %predicted | 48 ± 28 | 8–124 | ||
FEV1/FVC, % | 44 ± 17 | 11–86 | ||
PEFR, %predicted | 67 ± 28 | 25–155 | ||
RV/TLC, % | 46 ± 11 | 24–70 | ||
Dl CO/Va, ml/min/mm Hg/L | 3.81 ± 1.23 | 1.51–7.94 |
Mean ± SD | Range | |||
---|---|---|---|---|
LAA% | 31.7 ± 15.1 | 4.2–70.6 | ||
T, mm | 1.5 ± 0.2 | 1.2–2.3 | ||
T/, mm/m | 1.2 ± 0.2 | 0.9–1.9 | ||
Ai, mm2 | 17.3 ± 7.6 | 3.8–39 | ||
Ai/BSA, mm2/m2 | 10.8 ± 4.7 | 2.3–24.3 | ||
Ao, mm2 | 47.0 ± 12.9 | 18.1–84.1 | ||
Ao/BSA, mm2/m2 | 29.4 ± 8.3 | 10.8–52.4 | ||
T/D ratio | 0.204 ± 0.034 | 0.144–0.289 | ||
WA%, % | 64.5 ± 7.9 | 49.4–82.2 |
None of the 20 asymptomatic smokers had chronic bronchitis (by definition) and 59% of the obstructed patients had chronic bronchitis. However, there were no significant differences between the patients with and without chronic bronchitis in terms of T, T/, Ai, Ai/BSA, Ao, Ao/BSA, WA%, and T/D ratio. There were also no significant relationships between smoking index and airway dimensions.
Table 3 shows the correlation coefficients (r values) for univariate regression analyses of FEV1%P and CT parameters. As expected, we observed an inverse relationship between LAA% and FEV1%P. The data for airway measurements suggest that the airway wall is thicker and the luminal area smaller in patients who have more severe airflow obstruction. However, there was no significant relationship between Ao and FEV1%P either before or after the correction for BSA. There were no correlations between LAA% and WA% or between LAA% and T/D ratio. Therefore we adopted WA% as a primary measurement of airway abnormality and LAA% as an index of emphysema.
r Values | p Value* | |||
---|---|---|---|---|
LAA% | −0.529 | < 0.001 | ||
T, mm | −0.199 | NS | ||
T/, mm/m | −0.298 | < 0.01 | ||
Ai, mm2 | 0.273 | < 0.05 | ||
Ai/BSA, mm2/m2 | 0.192 | NS | ||
Ao, mm2 | 0.195 | NS | ||
Ao/BSA, mm2/m2 | 0.065 | NS | ||
T/D ratio | −0.333 | < 0.001 | ||
WA%, % | −0.338 | < 0.001 |
An average of 2.6 additional airways per case (range of 1 to 6) was measured in the subset of 20 subjects. The mean luminal area of the additional airways was 9.6 mm2 (range 2.4 to 24.7 mm2). The relationship between the WA% in the apical segmental bronchus and the average WA% in the additional airways was highly significant (r = 0.860, p < 0.0001, Figure 3).

Fig. 3. Relationship between WA% in the right apical bronchus and average WA% in the additional airways. There was a significant relationship between these two measurements (average WA% in the additional airways = 0.57 × WA% in the right apical bronchus + 36.7, r = 0.860, p < 0.0001).
[More] [Minimize]There were 37 subjects in the COPD ⩽ 30 group, 57 in the COPD > 30 group, and 20 in the Asymptomatics. The mean ± SD of WA% was 69 ± 8 in COPD ⩽ 30, 63 ± 7 in COPD > 30, and 59 ± 5 in Asymptomatics. There were significant differences between COPD ⩽ 30 and COPD > 30 (p < 0.0001), between COPD ⩽ 30 and Asymptomatics (p < 0.0001), and between COPD > 30 and Asymptomatics (p < 0.05). Figures 4A and 4B show the relationships for FEV1%P versus LAA% and FEV1%P versus WA%. COPD ⩽ 30, COPD > 30, and Asymptomatics are identified by separate symbols. There were significant negative relationships for FEV1%P and LAA% and for FEV1%P and WA%. These data show that emphysema and airway wall thickening increased as FEV1%P decreased.

Fig. 4. Relationship for FEV1 (percentage of predicted) versus LAA% (A) and versus WA% (B) in 114 smokers. FEV1 correlated negatively with both LAA% (LAA% = −0.29 × FEV1 + 45.7, r = −0.529, p < 0.001) and WA% (WA% = −0.095 × FEV1 + 69.1, r = −0.338, p < 0.001). Solid diamonds: COPD patients with the LAA% more than 30% (COPD > 30); open circles: COPD patients with the LAA% less than 30% (COPD ⩽ 30), open triangles: asymptomatic smokers (Asymptomatics).
[More] [Minimize]Table 4 shows univariate and stepwise multiple regression analyses of LAA% and WA% versus measurements of lung function. In the univariate regression analysis LAA% correlated significantly with FEV1%P, FEV1/FVC, PEFR%P, RV/ TLC, and Dl CO/Va but not FVC%P, whereas WA% correlated significantly with FVC%P, FEV1%P, PEFR%P, and RV/TLC but not FEV1/FVC or Dl CO/Va. It is also apparent from univariate regression analyses that LAA% correlates more closely with FEV1%P and FEV1/FVC, whereas WA% correlates more closely with PEFR%P and RV/TLC. To evaluate whether the measurements of airway dimensions by CT would provide additive value to the measurements of LAA in the estimate of lung function tests, we compared the results of univariate and stepwise multiple regression analysis (Table 4). The results showed that FVC%P, FEV1%P, FEV1/FVC, PEFR%P, and RV/TLC were predicted better by a combination of LAA% and WA% than by LAA% or WA% alone. However, WA% added no predictive value to LAA% for Dl CO/Va.
Univariate Regression Analysis | Multiple Regression Analysis LAA% and WA% | |||||
---|---|---|---|---|---|---|
LAA% | WA% | |||||
FVC, % predicted | −0.159† | −0.437* | 0.482* | |||
FEV1, % predicted | −0.529* | −0.338* | 0.659* | |||
FEV1/FVC, % | −0.650* | −0.192† | 0.700* | |||
PEFR, % predicted | −0.395* | −0.487* | 0.660* | |||
RV/TLC, % | 0.378* | 0.422* | 0.597* | |||
Dl CO/Va, ml/min/mm Hg/L | −0.683* | 0.030† | NA |
The measurement of in vivo airway dimensions using CT has been reported in animal models (24, 25) and in human subjects (19-23, 27). In most studies HRCT images and a visual assessment method were used (19-24). The visual assessment method is time-consuming (22) and intraobserver or interobserver error is not trivial. We used a computer-based method to measure airway dimensions. This technique is faster and more reproducible. In the present study, we used thin-section helical CT images to assess the airway dimensions because the contiguous images allowed more accurate identification of the right apical segmental bronchus. It has been reported that HRCT can accurately assess the dimensions of hollow tubes of 1 to 5 mm in diameter with a thickness from 0.5 to 2 mm (24, 30). We used the “full-width half-maximum” technique (25, 26) to estimate the wall thickness (see ) and used a single threshold (−500 HU) (27) to measure the luminal area. It has been reported that this method is suitable for airways with a large inside diameter and thick walls but can produce an underestimate of luminal area and an overestimate of wall thickness in small airways and airways with thin walls (26). Therefore, we chose to measure a relatively large and thick-walled segmental bronchus. The validation study of the airway phantoms using the thin-section helical CT technique showed that our system and algorithm are accurate within the range of 1.0 to 2.3 mm for wall thickness and 12 to 335 mm2 for luminal area. However, our method overestimates the thickness when it is less than 1.0 mm and underestimates the luminal area when it is less than 12 mm2. Because the ranges of the airway measurements in this study were from 1.2 to 2.3 mm for thickness and from 3.8 to 39.0 mm2 for luminal area, underestimation of the luminal area may affect the result. However, when we limited the analysis to only those subjects whose luminal area was greater than 12 mm2 (86/114 subjects), the results of the analysis were unchanged.
We measured a single specific airway in the present study because measuring airway wall dimensions without respect to standardization of anatomic location could present problems; specifically wall area as a percent of wall and lumen area varies by anatomic location in the tracheobronchial tree (31). Further, our aim was to test whether a relatively simple measurement that could be added to conventional CT scanning would be of predictive value for airflow obstruction. To test the generalizibility of the measurements on this specific airway, we randomly chose 20 subjects from the total group and measured WA% on all other airways that were sectioned in cross-section in the helical slices to compare the average WA% in airways other than the apical segmental bronchus with the apical segmental bronchus. The highly significant relationship between WA% in the apical segmental bronchus and the additional airways (Figure 3) indicates that the dimensions of this airway are representative of other airways in the same subject. However, the intercept was not zero and the slope was not unity. The reason that the intercept is positive is likely due to the fact that smaller airways have a relatively greater airway wall area (31). We cannot explain why the slope is less than one. One possibility is that airway walls can only thicken to a certain extent. There appears to be a maximal WA% of approximately 80. If this is the case, this could explain the slope.
Our finding of thickened cartilaginous airways in smokers is consistent with the results of other investigators. Haraguchi and colleagues (32) showed that the bronchi of patients with COPD had more degenerated cartilage and perichondrial fibrosis than a control group. They also showed that the degree of perichondrial fibrosis correlated with the thickness of the epithelial basement membrane. Tiddens and colleagues (33) evaluated airway wall dimensions in cartilaginous airways of patients with COPD. They found that the wall area of central airways was increased in obstructed patients (FEV1/FVC = 40%) in comparison to nonobstructed patients (FEV1/FVC = 80%). They also showed that the increase in cartilaginous airway wall area correlated significantly with a semiquantitative measure of peripheral airway inflammation. It has also been shown that the small airways of patients with mild COPD have thicker walls and narrower lumens than control subjects (31, 34). Because we made only a single airway measurement in most subjects, we cannot comment on the heterogeneity of changes in airway dimensions. However, such heterogeneity would decrease the chances of finding a relationship between physiologic measurements and morphometric measurements on a single airway. The finding of a significant correlation between the airway dimensions and measurements of airflow obstruction suggests that large airway dimensions may reflect small airway pathology.
Bosken and coworkers (34) found that smoking history was unrelated to the morphometric measures of peripheral airway dimensions. In the present study, there were not significant relationships between smoking index and airway dimensions.
Our data show that airway wall area and thickness were greater and airway luminal area was smaller as FEV1%P decreased. In contrast, there was no significant relationship between Ao and FEV1%P. These results suggest that as the airway thickens, it encroaches upon the lumen rather than expanding outward into the lung parenchyma. Alternatively, the airway wall thickening could be accompanied by a degree of airway smooth muscle shortening which could have the same effect.
Airflow obstruction in smokers is caused by loss of lung recoil and airway narrowing. In any one subject it is difficult to define the relative contributions of these mechanisms to the airflow obstruction. Emphysema, and its CT surrogate LAA%, correlates with loss of recoil but to date there have been no objective measurements of airway narrowing by CT in smokers. Our results show that measures of large airway thickening correlate with abnormalities of lung function except FEV1/FVC and Dl CO/Va. Wall thickening correlated especially well with FVC%P and RV/TLC, suggesting that airway wall thickening is associated with gas trapping. However, because RV was measured using a helium dilution method, which may lead to the underestimation of RV in severely obstructed individuals, careful interpretation is required. The correlation of wall area and peak expiratory flow suggests that wall thickening influences maximal flow early in forced expiration when peak flow occurs. The correlation of WA% and FEV1%P also suggests that wall thickening and airway narrowing of large airways influences airflow obstruction. One possibility is that large airway wall thickening is simply an indicator of more peripheral airway inflammation and narrowing because this is the site of the major obstruction in smokers (35). However, the finding that WA% correlates most closely with PEFR%P, a measure known to be influenced by large airway caliber, suggests a direct effect of large airway remodeling on expiratory flow.
The results of the stepwise multiple regression analysis suggest that WA% and LAA% are measuring independent aspects of pulmonary pathophysiology in smokers: the airway and parenchymal components, respectively. There were significant differences in WA% between COPD ⩽ 30 and COPD > 30, between COPD ⩽ 30 and Asymptomatics, and between COPD > 30 and Asymptomatics. These results suggest that at comparable levels of FEV1, the patients with more emphysema have less severe airway thickening than those with less severe emphysema. However, even those with more severe emphysema had thicker airways than the asymptomatic smokers.
In summary, the novel finding in this study is that measurements of airway dimensions in a single airway using a protocol that can be easily added to CT assessment of emphysema yield additive information about the mechanism of airflow obstruction in smokers.
The authors thank Drs. M. Okazawa, A. Niimi, H. O. Coxson, and J. R. Mayo for useful suggestions. They also thank Mr. R. Tanaka, Ms. M. Morimoto, Mr. H. Akazawa, and Mr. N. Narai for their technical assistance with CT.
Supported by the Research Council of Respiratory Insufficiency, Ministry of Public Health, Japan, and the Ministry of Education, Science, Sports and Culture, Japan.
1. | ATS StatementStandards for the diagnosis and care of patients with chronic obstructive pulmonary disease. Am. J. Respir. Crit. Care Med.152(Suppl.)1995S77S83 |
2. | Bergin C., Müller N., Nichols D. M., Lillington G., Hogg J. C., Mullen B., Grymaloski M. R., Osborne S., Paré P. D.The diagnosis of emphysema: a computed tomographic–pathologic correlation. Am. Rev. Respir. Dis.1331986541546 |
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. | 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 |
5. | Gevenois P. A., De Vuyst P., de Maertelaer V., Zanen J., Jacobovitz D., Cosio M. G., Yernault J.-C.Comparison of computed density and microscopic morphometry in pulmonary emphysema. Am. J. Respir. Crit. Care Med.1541996187192 |
6. | Gevenois P. A., De Vuyst P., Sy M., Scillia P., Chaminade L., de Maertelaer V., Zanen J., Yernault J.-C.Pulmonary emphysema: quantitative CT during expiration. Radiology1991996825829 |
7. | Kuwano K., Matsuba K., Ikeda T., Murakami J., Araki A., Nishitani H., Ishida T., Yasumoto K., Shigematsu N.The diagnosis of mild emphysema: correlation of computed tomography and pathology scores. Am. Rev. Respir. Dis.1411990169178 |
8. | Sakai N., Mishima M., Nishimura K., Itoh H., Kuno K.An automated method to assess the distribution of low attenuation areas on chest CT scans in chronic pulmonary emphysema patients. Chest106199413191325 |
9. | Lamers R. J., Thelissen G. R., Kessels A. G., Wouters E. F., van Engelshoven J. M.Chronic obstructive pulmonary disease: evaluation with spirometrically controlled CT lung densitometry. Radiology1931994109113 |
10. | Gurney J. W., Jones K. K., Robbins R. A., Gossman G. L., Nelson K. J., Daughton D., Spurzem J. R., Rennard S. I.Regional distribution of emphysema: correlation of high-resolution CT with pulmonary function tests in unselected smokers. Radiology1831992457463 |
11. | Kinsella M., Müller N. L., Abboud R. T., Morrison N. J., DyBuncio A.Quantitation of emphysema by computed tomography using a “density mask” program and correlation with pulmonary function tests. Chest971990315321 |
12. | Coxson H. O., Rogers R. M., Whittall K. P., D'Yachkova Y., Paré P. D., Sciurba F. C., Hogg J. C.A quantification of the lung surface area in emphysema using computed tomography. Am. J. Respir. Crit. Care Med.1591999851856 |
13. | Nishimura K., Murata K., Yamagishi M., Itoh H., Ikeda A., Tsukino M., Koyama H., Sakai N., Mishima M., Izumi T.Comparison of different computed tomography scanning methods for quantifying emphysema. J. Thoracic Imaging131998193198 |
14. | Nakano Y., Sakai H., Muro S., Hirai H., Oku Y., Nishimura K., Mishima M.Comparison of low attenuation areas on CT between inner and outer segments of the lung in COPD patients: incidence and contribution to lung function. Thorax541999384389 |
15. | Mishima M., Hirai T., Itoh H., Nakano Y., Sakai H., Muro S., Nishimura K., Oku Y., Chin K., Ohi M., Nakamura T., Bates J. H. T., Alencar A. M., Suki B.Complexity of terminal airspace geometry assessed by lung CT in normal subjects and patients with chronic obstructive pulmonary disease. Proc. Natl. Acad. Sci. U.S.A.96199988298834 |
16. | Mishima M., Oku Y., Kawakami K., Sakai N., Fukui M., Hirai T., Chin K., Ohi M., Nishimura K., Itoh H., Tanemura M., Kuno K.Quantitative assessment of the spatial distribution of low attenuation areas on X-ray CT using texture analysis in patients with chronic pulmonary emphysema. Front. Med. Biol. Eng.819971934 |
17. | Mishima M., Itoh H., Sakai H., Nakano Y., Muro S., Hirai T., Takubo Y., Chin K., Ohi M., Nishimura K., Yamaguchi K., Nakamura T.Optimized scanning conditions of HRCT in the follow-up of pulmonary emphysema. J. Comput. Assist. Tomogr.231999380384 |
18. | Mishima M., Hirai T., Jin Z., Oku Y., Sakai N., Nakano Y., Sakai H., Chin K., Ohi M., Kawakami K., Shimada K., Itoh H., Yamaguchi K., Sawa T., Kuno K.Standardization of low attenuation area versus total lung area in chest X-ray CT as an indicator of chronic pulmonary emphysema. Front. Med. Biol. Eng.819977986 |
19. | Okazawa M., Müller N., McNamara A. E., Child S., Verburgt L., Paré P. D.Human airway narrowing measured using high resolution computed tomography. Am. J. Respir. Crit. Care Med.154199615571562 |
20. | Boulet L.-P., Turcotte H., Carrier G., Boutet M., Laviolette M.Increased maximal airway response to methacholine during seasonal allergic rhinitis in nonasthmatic subjects: relationships with airway wall thickness and inflammation. Eur. Respir. J.81995913921 |
21. | Boulet L.-P., Bélanger M., Carrier G.Airway responsiveness and bronchial-wall thickness in asthma with or without fixed airflow obstruction. Am. J. Respir. Crit. Care Med.1521995865871 |
22. | Awadh N., Müller N. L., Park C. S., Abboud R. T., FitzGerald J. M.Airway wall thickness in patients with near fatal asthma and control groups: assessment with high resolution computed tomographic scanning. Thorax531998248253 |
23. | Senéterre E., Paganin F., Bruel J. M., Michel F. B., Bousquet J.Measurement of the internal size of bronchi using high resolution computed tomography (HRCT). Eur. Respir. J.71994596600 |
24. | McNamara A. E., Müller N. L., Okazawa M., Arntorp J., Wiggs B. R., Paré P. D.Airway narrowing in excised canine lungs measured by high-resolution computed tomography. J. Appl. Physiol.731992307316 |
25. | Amirav I., Kramer S. S., Grunstein M. M., Hoffman E. A.Assessment of methacholine-induced airway constriction by ultrafast high-resolution computed tomography. J. Appl. Physiol.75199322392250 |
26. | Reinhardt J. M., D'Souza N. D., Hoffman E. A.Accurate measurement of intrathoracic airways. IEEE Trans. Med. Imaging161997820827 |
27. | McNitt-Gray M. F., Goldin J. G., Johnson T. D., Tashkin D. P., Aberle D. R.Development and testing of image-processing methods for the quantitative assessment of airway hyperresponsiveness from high-resolution CT images. J. Comput. Assist. Tomogr.211997939947 |
28. | King G. G., Müller N. L., Paré P. D.Evaluation of airways in obstructive pulmonary disease using high-resolution computed tomography. Am. J. Respir. Crit. Care Med.15919999921004 |
29. | Japanese Society of Chest Diseases. 1993. Standards of pulmonary function tests for Japanese. Jpn. J. Thorac. Dis. 31:appendix. |
30. | Webb W. R., Gamsu G., Wall S. D., Cann C. E., Proctor E.CT of a bronchial phantom: factors affecting appearance and size measurements. Invest. Radiol.191984394398 |
31. | Kuwano K., Bosken C. H., Paré P. D., Bai T. R., Wiggs B. R., Hogg J. C.Small airways dimensions in asthma and in chronic obstructive pulmonary disease. Am. Rev. Respir. Dis.148199312201225 |
32. | Haraguchi M., Shimura S., Shirato K.Morphometric analysis of bronchial cartilage in chronic obstructive pulmonary disease and bronchial asthma. Am. J. Respir. Crit. Care Med.159199910051013 |
33. | Tiddens H. A. W. M., Paré P. D., Hogg J. C., Hop W. C. J., Lambert R., de Jongste J. C.Cartilaginous airway dimensions and airflow obstruction in human lungs. Am. J. Respir. Crit. Care Med.1521995260266 |
34. | Bosken C. H., Wiggs B. R., Paré P. D., Hogg J. C.Small airway dimensions in smokers with obstruction to airflow. Am. Rev. Respir. Dis.1421990563570 |
35. | 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 |