A semiautomatic method of quantifying pleural effusions using computed tomography scans
Pleural effusion (PE) is a common and devastating manifestation of many pleural diseases. This complication can have dire effects by constricting breathing which increases the recovery time of the patient. Increased volumes of PE correspond with increased complications from the disorder. For this study, accurate PE volumes were found by segmentation of Computed Tomography (CT) scans. The retrospective study, with each scan conducted out of medical necessity for patient care or diagnosis, used a dataset of 105 CT scans and 179 individual pleural effusions. A semiautomatic method of segmentation was developed to quantify the volume of the PE. For this method, a thresholding method based on Hounsfield unit values of the tissues was used to determine six volumes (total lung, total pleural effusion, right lung, right pleural effusion, left lung and left pleural effusion). The volume of each PE was divided by the total lung volume to calculate PE as a percentage of the total lung. This normalization allows for direct comparison between each subject. The segmentations were conducted for a wide range of pleural effusion severities with a range from 0.122 to 67.798 percent of total lung volume. Most effusions were small with an average volume of 6.66 ± 12.22 percent of total lung volume. These data will be used in a future study investigating the relationship between segmented PE volume and PE volume calculated from 2D chest images. © 2013.
Anderson, SA; Danelson, KA; Mammarappallil, JG; Chiles, C; Stitzel, JD
Volume / Issue
Start / End Page
International Standard Serial Number (ISSN)