Skip to Main Content
Principal Investigator
David Gierada
Washington University
Position Title
Professor of Radiology
About this CDAS Project
NLST (Learn more about this study)
Project ID
Initial CDAS Request Approval
Jul 30, 2004
To Determine the Effects of Slice Thickness and Reconstruction Filter on Measurements of Lung Attenuation Obtained from Low Dose Multidetector CT Scans
The purpose of this study is to determine the effects of slice thickness and reconstruction filter on measurements of lung attenuation obtained from low dose multidetector CT scans. Previous studies have validated and used CT measurements of emphysema based on lung attenuation, but have not fully evaluated the impact of these variables. The overall goal is to devise methods for comparing emphysema measurements on scans that were reconstructed from multidetector CT scans using different techniques. To perform this study, we will reconstruct the raw data from Washington University NLST CT scans at multiple slice thicknesses, with multiple reconstruction filters. The scans will be analyzed on a desktop computer to obtain histograms of lung pixel attenuation values and calculate the percentage of the lung having attenuation values below multiple thresholds (attenuation indexes). The relationships between attenuation indexes obtained with different slice thicknesses and reconstruction filters will be defined. Regression analyses will be performed in an attempt to determine conversion factors that may allow measurements obtained with one reconstruction technique to be compared with those obtained using a different technique. The impact of the different reconstruction techniques on the diagnosis of emphysema and the ability to assess emphysema progression using different reconstruction techniques will be evaluated. We will use only NLST scan data that has already been obtained, and no additional consent from the NLST participants will be required. The scans will be used only for attenuation histogram analyses, and there will be no clinically relevant results to report. The NLST screening results and other NLST data will not be used.

CT measurements of lung attenuation have been shown to correlate with the amount of emphysema. Using CT, emphysema is usually quantified as the percentage of the lung having an attenuation below a specific threshold value. Studies using single detector scanners have found that the optimal attenuation threshold for defining emphysema depends on the slice thickness (Muller et al, 1988; Gevenois et al, 1995). Lung attenuation values also vary depending on the reconstruction algorithm applied to the raw data (Boedecker et al, 2004s). Thus, quantitative CT measurements of emphysema may be significantly influenced by how a study was reconstructed. The purpose of this study is to define the relationships between the different reconstruction algorithms and attenuation values in lung tissue on multidetector CT scans. The results in turn may be useful for comparing emphysema measurements on scans obtained using different reconstruction methods. In this study, we will analyze and compare NLST CT scans using multiple slice thicknesses and reconstruction filters in order to: 1. Understand the impact of different reconstruction algorithms (i.e., slice thickness and reconstruction filter) on the attenuation values of lung tissue. 2. Develop a regression model that predicts lung attenuation values from various slice thicknesses and reconstruction filters, thereby determining conversion factors for comparing emphysema measurements obtained using different reconstruction parameters. 3. Utilize the relationships between the reconstruction parameters and various attenuation thresholds for defining emphysema, to determine how scanner reconstruction algorithms may impact the classification of emphysema severity. 4. Determine the prevalence of emphysema in the Washington University NLST population.

Related Publications