Quantitative assessment of emphysema by morphological CT measurements in COPD patients.

Jansen, S. (Sanne) (2015) Quantitative assessment of emphysema by morphological CT measurements in COPD patients. thesis, Medicine.

Text JansenS.pdf Restricted to Registered users only Download (905kB)

Abstract

Introduction Worldwide 65 million people suffer from COPD (chronic obstructive pulmonary disease) and it is expected to be the third cause of death in 2030. At this moment this complex disease is not completely understood and therefore a better understanding is desired. Post-processing of CT scans may contribute to a better insight. It is a measure to calculate the percentage of emphysema and this radiological measure may correlate with the signs and symptoms of COPD. The purpose of this study is to determine the optimal level of reconstruction for post-processing emphysema quantification in COPD patients. In addition correlations between pulmonary function tests and the percentage of emphysema will be determined. Furthermore the estimations of a pulmonologist are compared to the post-processing software and the workflow and usability of post-processing are analyzed. Material and Methods 28 patients with COPD who received a CT scan and a pulmonary function test less than 6 months apart were included. Also, the MRC scale of dyspnea and the COPD Control Questionnaire (CCQ) to measure health status were administered. The pulmonary function test parameters chosen were the FEV1% predicted and FEV1/FVC ratio. Standing for the ratio of the forced expiratory volume and the forced vital capacity which has to be <0.7 for the classification of COPD. A study-specific CT application form was used to assure all patients received the same scan. For each patient all iterative reconstruction levels were reconstructed. So for each patient seven emphysema percentages were calculated with the IntelliSpace Portal. The emphysema parameter used was %LAA-950, the percentage of lung attenuation areas under -950HU, an established measure for emphysema on CT scans. The optimal level of iterative reconstruction is determined through regression and correlation. The pulmonary function parameters were tested with each level separately. The data was also tested on significant differences between the iDose levels with a Friedman test. The correlations were tested with Pearson for parametric data and Spearman for nonparametric data. The agreement between the estimations of the pulmonologist and the software was tested with Cohens Kappa. A short questionnaire was sent to the pulmonologists about the usability of the post-processing information. Results Through univariate linear regression the highest association was found between level 6 and FEV1/FVC with a correlation coeffictient of -0.702 and a regression coefficient of -0.160 (p-value<0.001). The correlation coefficient found appeared to be similar to values from literature. Friedman test found a significant difference between the iDose levels, post hoc Wilcoxon test with a Bonferroni adjustment were performed to locate the differences. All combinations of levels turned out significant (p-value<0.001). There was no association found with FEV1% predicted. The estimations of the pulmonologist and the software had no significant difference with a Kappa of 0,383 and a p-value< 0.001. The workflow concerning the calculation of the emphysema percentage of the pulmonologist was similar to the software but the software offers more information but it is not clear yet how the information could be used in de daily practice. Conclusion In this population iDose level 6 seems to be most optimal for the quantification of emphysema in patients with COPD. Another important finding is the significant difference between the iDose levels, so caution should be taken when choosing an iDose level. Furthermore correlations corresponding with literature were found and it appears that pulmonologists do not underestimate the percentage of emphysema. The use of post-processing in the daily practice must explored further but will not be stopped by a longer workflow.

Item Type:	Thesis (Thesis)
Supervisor name:	Berg, dr. J.W.K. van den and Pulmonology, Radiology and Isala Klinieken Zwolle
Faculty:	Medical Sciences
Date Deposited:	25 Jun 2020 10:58
Last Modified:	25 Jun 2020 10:58
URI:	https://umcg.studenttheses.ub.rug.nl/id/eprint/1898

Actions (login required)

View Item