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

This Article Abstract Full Text Full Text (PDF) 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 PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Berger, P. Articles by Laurent, F. Search for Related Content PubMed PubMed Citation Articles by Berger, P. Articles by Laurent, F.

Airway Wall Thickness in Cigarette Smokers: Quantitative Thin-Section CT Assessment¹

¹ From the Laboratoire de Physiologie Cellulaire Respiratoire (Institut National de la Santé et de la Recherche Médicale E-0356), Université Victor Ségalen, Bordeaux, France (P.B., J.M.T.d.L., R.M., F.L.); and Unité d’Imagerie Thoracique et Cardiovasculaire, Hôpital Cardiologique Haut-Lévêque, Avenue de Magellan, 33604 Pessac, France (V.P., P.D., F.L.). Received January 26, 2004; revision requested April 2; final revision received August 12; accepted September 15. Supported by grants from Programme Hospitalier de Recherche Clinique in 1997 and 2002. Address correspondence to F.L. (e-mail: francois.laurent@chu-bordeaux.fr).

View larger version (71K): [in a new window]   Figure 1. Laplacian of Gaussian algorithm. A, Diagram shows region of interest symbolized by a circle. B, Graph shows analysis of region of interest with a gray-level profile of a line in the middle of the circle. C, Graph shows first derivative of the profile (gradient). D, Graph shows second derivative (Laplacian). E-H, Transverse sections show various steps of segmentation of two bronchi in a 57-year-old healthy male control subject. Images before (E) and after (F) five times magnification depict tracings of square regions of interest that encompassed the selected bronchi. G, Image shows the calculation of the Laplacian of Gaussian mask that was performed with the algorithm and application of a spatial convolution to the whole image. H, Image shows selected region of interest magnified by a factor of five and the resulting binary image that was edited to separate the bronchus wall from adjacent vessels and to suppress the internal pixels caused by noise.

View larger version (147K): [in a new window]   Figure 2. Images from inflated fixed excised lung study. A, Digitized image of the cut surface of transverse lung slice with a square region of interest. B, Corresponding thin-section CT image with same square region of interest. C, Image shows airway cross section of interest, with, D, internal and external contours. E, Thin-section CT image can be matched with corresponding image in C after semiautomatic segmentation and, F, filtering. IA and WA measurements can be automatically calculated.

View larger version (19K): [in a new window]   Figure 3a. Graphs of data from inflated fixed excised lung study. Measurements on CT scans were plotted against the mean measurements performed manually on fixed sections by two independent observers. The diagonal line corresponds to the line of equality. There was a strong correlation between DACLOG data and those obtained by both observers as assessed with the intraclass correlation coefficient (ICC). (a) DACLOG IA measurements. (b) DACLOG WA measurements.

View larger version (19K): [in a new window]   Figure 3b. Graphs of data from inflated fixed excised lung study. Measurements on CT scans were plotted against the mean measurements performed manually on fixed sections by two independent observers. The diagonal line corresponds to the line of equality. There was a strong correlation between DACLOG data and those obtained by both observers as assessed with the intraclass correlation coefficient (ICC). (a) DACLOG IA measurements. (b) DACLOG WA measurements.

View larger version (20K): [in a new window]   Figure 4a. Graphs of data from inflated fixed excised lung study. Means of measurements obtained with DACLOG and manual method are plotted against their differences according to Bland-Altman analysis. Solid line corresponds to the mean difference, and dashed lines, to the mean difference ± 2 standard deviations. The lack of agreement was greater for WA than it was for IA. (a) IA measurements. (b) WA measurements.

View larger version (21K): [in a new window]   Figure 4b. Graphs of data from inflated fixed excised lung study. Means of measurements obtained with DACLOG and manual method are plotted against their differences according to Bland-Altman analysis. Solid line corresponds to the mean difference, and dashed lines, to the mean difference ± 2 standard deviations. The lack of agreement was greater for WA than it was for IA. (a) IA measurements. (b) WA measurements.

View larger version (20K): [in a new window]   Figure 5a. Graphs of data from inflated fixed excised lung study. Means of DACLOG measurements are plotted against standard deviations of DACLOG measurements performed by two independent observers at different times. The coefficient r corresponds to the Pearson correlation coefficient, which shows correlation between standard deviation and the mean of these measurements. Standard deviations of IA and WA assessed with DACLOG were not correlated with the mean values of IA and WA, respectively. NS = not significant. (a) IA measurements. (b) WA measurements.

View larger version (20K): [in a new window]   Figure 5b. Graphs of data from inflated fixed excised lung study. Means of DACLOG measurements are plotted against standard deviations of DACLOG measurements performed by two independent observers at different times. The coefficient r corresponds to the Pearson correlation coefficient, which shows correlation between standard deviation and the mean of these measurements. Standard deviations of IA and WA assessed with DACLOG were not correlated with the mean values of IA and WA, respectively. NS = not significant. (a) IA measurements. (b) WA measurements.

View larger version (16K): [in a new window]   Figure 6a. Box plots show distribution of parameters in 970 bronchi by using DACLOG. Group 1 corresponds to nonsmokers with normal lung function; group 2, smokers with normal lung function; and group 3, smokers with COPD. (a-d) Medians with 25% and 75% interquartiles; error bars represent 5th and 95th percentiles. (e, f) Bars represent the mean, and error bars represent the standard error of the mean for each group. The parameters that reflect airway thickening (WA/IA and WA/[IA + WA] ratios) could be used to discriminate smokers from nonsmokers even in the absence of COPD. (a) IA + WA. (b) IA. (c) WA. (d) WA/IA ratio. (e) Individual WA/IA ratio. (f) Individual WA/(IA + WA) ratio.

View larger version (15K): [in a new window]   Figure 6b. Box plots show distribution of parameters in 970 bronchi by using DACLOG. Group 1 corresponds to nonsmokers with normal lung function; group 2, smokers with normal lung function; and group 3, smokers with COPD. (a-d) Medians with 25% and 75% interquartiles; error bars represent 5th and 95th percentiles. (e, f) Bars represent the mean, and error bars represent the standard error of the mean for each group. The parameters that reflect airway thickening (WA/IA and WA/[IA + WA] ratios) could be used to discriminate smokers from nonsmokers even in the absence of COPD. (a) IA + WA. (b) IA. (c) WA. (d) WA/IA ratio. (e) Individual WA/IA ratio. (f) Individual WA/(IA + WA) ratio.

View larger version (16K): [in a new window]   Figure 6c. Box plots show distribution of parameters in 970 bronchi by using DACLOG. Group 1 corresponds to nonsmokers with normal lung function; group 2, smokers with normal lung function; and group 3, smokers with COPD. (a-d) Medians with 25% and 75% interquartiles; error bars represent 5th and 95th percentiles. (e, f) Bars represent the mean, and error bars represent the standard error of the mean for each group. The parameters that reflect airway thickening (WA/IA and WA/[IA + WA] ratios) could be used to discriminate smokers from nonsmokers even in the absence of COPD. (a) IA + WA. (b) IA. (c) WA. (d) WA/IA ratio. (e) Individual WA/IA ratio. (f) Individual WA/(IA + WA) ratio.

View larger version (15K): [in a new window]   Figure 6d. Box plots show distribution of parameters in 970 bronchi by using DACLOG. Group 1 corresponds to nonsmokers with normal lung function; group 2, smokers with normal lung function; and group 3, smokers with COPD. (a-d) Medians with 25% and 75% interquartiles; error bars represent 5th and 95th percentiles. (e, f) Bars represent the mean, and error bars represent the standard error of the mean for each group. The parameters that reflect airway thickening (WA/IA and WA/[IA + WA] ratios) could be used to discriminate smokers from nonsmokers even in the absence of COPD. (a) IA + WA. (b) IA. (c) WA. (d) WA/IA ratio. (e) Individual WA/IA ratio. (f) Individual WA/(IA + WA) ratio.

View larger version (23K): [in a new window]   Figure 6e. Box plots show distribution of parameters in 970 bronchi by using DACLOG. Group 1 corresponds to nonsmokers with normal lung function; group 2, smokers with normal lung function; and group 3, smokers with COPD. (a-d) Medians with 25% and 75% interquartiles; error bars represent 5th and 95th percentiles. (e, f) Bars represent the mean, and error bars represent the standard error of the mean for each group. The parameters that reflect airway thickening (WA/IA and WA/[IA + WA] ratios) could be used to discriminate smokers from nonsmokers even in the absence of COPD. (a) IA + WA. (b) IA. (c) WA. (d) WA/IA ratio. (e) Individual WA/IA ratio. (f) Individual WA/(IA + WA) ratio.

View larger version (24K): [in a new window]   Figure 6f. Box plots show distribution of parameters in 970 bronchi by using DACLOG. Group 1 corresponds to nonsmokers with normal lung function; group 2, smokers with normal lung function; and group 3, smokers with COPD. (a-d) Medians with 25% and 75% interquartiles; error bars represent 5th and 95th percentiles. (e, f) Bars represent the mean, and error bars represent the standard error of the mean for each group. The parameters that reflect airway thickening (WA/IA and WA/[IA + WA] ratios) could be used to discriminate smokers from nonsmokers even in the absence of COPD. (a) IA + WA. (b) IA. (c) WA. (d) WA/IA ratio. (e) Individual WA/IA ratio. (f) Individual WA/(IA + WA) ratio.

View larger version (189K): [in a new window]   Figure 7a. Representative transverse CT images. Squares were traced around the bronchi chosen for analysis by using DACLOG. By using a qualitative approach, it was difficult to differentiate smokers from nonsmokers even in the absence of COPD. (a) Male 51-year-old nonsmoker with normal lung function from group 1. (b) Female 53-year-old smoker with normal lung function from group 2. (c) Male 61-year-old smoker with COPD from group 3.

View larger version (173K): [in a new window]   Figure 7b. Representative transverse CT images. Squares were traced around the bronchi chosen for analysis by using DACLOG. By using a qualitative approach, it was difficult to differentiate smokers from nonsmokers even in the absence of COPD. (a) Male 51-year-old nonsmoker with normal lung function from group 1. (b) Female 53-year-old smoker with normal lung function from group 2. (c) Male 61-year-old smoker with COPD from group 3.

View larger version (175K): [in a new window]   Figure 7c. Representative transverse CT images. Squares were traced around the bronchi chosen for analysis by using DACLOG. By using a qualitative approach, it was difficult to differentiate smokers from nonsmokers even in the absence of COPD. (a) Male 51-year-old nonsmoker with normal lung function from group 1. (b) Female 53-year-old smoker with normal lung function from group 2. (c) Male 61-year-old smoker with COPD from group 3.