Comparing clinical variables and emphysema quantification using different reconstruction methods on HRCT scans in COPD patients

Ismailova, G. (Gamida) (2017) Comparing clinical variables and emphysema quantification using different reconstruction methods on HRCT scans in COPD patients. thesis, Medicine.

Text IsmailovaG.pdf Restricted to Registered users only Download (1MB)

Abstract

troduction COPD is a common disease that is characterized by respiratory symptoms and airflow limitation. It comprises the fourth cause of deaths worldwide. Inflammation in the airways leads to combinations of airway remodelling and emphysema, resulting in narrowed airways, airflow limitation, hyperinflation and reduced effective gas exchange of the lungs. Thus dyspnoea, reduced physical activity and health status occur. Currently, an increasing correlation between airflow limitation and emphysema and a decreasing amount of quantified emphysema is shown after the use of advanced iterative reconstruction methods on HRCT-scans. These methods reduce image noise and improve image quality, with an expectation of approaching the true value of emphysema even more. For a long time Filtered Back Projection (FBP) has been used as the standard reconstruction method for CT-scans. More advanced iterative reconstruction methods, such as iDose4, using statistics for noise reduction, have improved image quality and shown some of the afore mentioned developments. However there are no recent studies comparing emphysema quantification by these two techniques in combination with multiple clinical variables, where attention is given to the quantification differences at distinct amounts of emphysema. A more true amount of emphysema and a stronger correlation with airflow limitation and other clinical variables is expected after the iterative reconstruction method iDose4, due to advanced images. The emphysema quantification difference between the methods is expected to be independent of the total amount of emphysema present. These developments will contribute to a better understanding of the COPD pathophysiology and a better use of diagnostic tools for emphysema detection. Methods 40 patients referred to the Pulmonology Department, received HRCT-scans and six minutes walking tests, after diagnosis of COPD based on GOLD criteria. Lung function measures included FEV1, FEV1/FVC, FRC, RV, TLC, RV/TLC and diffusing capacity TLco. Health related quality of life measures consisted of the CCQ and MRC questionnaires. Emphysema percentages were quantified after reconstruction methods FBP and iDose4 level 6. Spearman’s correlation coefficients were determined between clinical variables and emphysema. Multiple binary logistic regression analysis was applied to determine whether the best correlated clinical variables were true predictors of emphysema. ROC curves were applied to determine the sensitivity of the best correlated clinical variable for detection of emphysema. A Wilcoxon signed-rank test stated the difference in emphysema quantification between the reconstruction methods. A sign test is used to determine whether this difference was dependent on different emphysema amounts. Results Emphysema percentages ranged between 0.1-37.7% for FBP and 0.2-38.2% for iDose4. The Wilcoxon signed-rank test presented a significant difference between these quantifications (p<.01), with 33 cases of lower quantified emphysema after iDose4. Emphysema and airflow limitation significantly correlated with values of -.614 (FBP) and -.591(iDose4). Emphysema and diffusing capacity correlated at best with values of -.686 (FBP) and -.680(iDose4). All significant correlations were lower at iDose4. Diffusing capacity showed 89% sensitivity for detection of emphysema at the cut-off value of 5% emphysema. iDose4 showed more quantified emphysema <10% compared to FBP. When dividing quantified emphysema in steps of 10% for each reconstruction method, this showed observed differences at values <30% and equality >30% of emphysema between iDose4 and FBP. However the sign test stated these differences non-significant. Conclusion iDose4 showed significantly different, mostly lower, quantified emphysema amounts compared to FBP. However, correlation coefficients between clinical variables and emphysema after iDose4 were not stronger. Observed differences in emphysema quantification existed between iDose4 and FBP at emphysema amounts <30% and an equal quantification existed at emphysema >30%, however without statistical significance. This means care must be taken when selecting a reconstruction method when quantifying emphysema. Based on our study, emphysema quantified after iterative reconstruction does not consequently generate a stronger association with lung function. Presence of emphysema could still be missed by the best correlated clinical variables, meaning a true amount of emphysema should still be approached with CT-scans with the use of an appropriate reconstruction method.

Item Type:	Thesis (Thesis)
Supervisor name:	Berg, Dr. J.W.K van den and Isala, Zwolle and Departments: and Pulmonology & and Radiology
Faculty:	Medical Sciences
Date Deposited:	25 Jun 2020 11:02
Last Modified:	25 Jun 2020 11:02
URI:	https://umcg.studenttheses.ub.rug.nl/id/eprint/2226

Actions (login required)

View Item