Improved correction for the tissue fraction effect in lung PET/CT imaging. (8th September 2015)

• Record Type: Journal Article
• Authors: Holman, Beverley F; Cuplov, Vesna; Millner, Lynn; Hutton, Brian F; Maher, Toby M; Groves, Ashley M; Thielemans, Kris
• Is Part Of: Physics in medicine & biology. Volume 60:Number 18(2015:Sep.)
• Page Start: 7387
• Page End: 7402
• Publication Date: 2015-09-08
• Subjects: lung -- PET/CT -- tissue fraction correction -- 18F-FDG -- idiopathic pulmonary fibrosis; Biophysics -- Periodicals; Medical physics -- Periodicals; 610.153
• DOI: 10.1088/0031-9155/60/18/7387
• Languages: English
• ISSNs: 0031-9155

Abstract

Abstract: Recently, there has been an increased interest in imaging different pulmonary disorders using PET techniques. Previous work has shown, for static PET/CT, that air content in the lung influences reconstructed image values and that it is vital to correct for this 'tissue fraction effect' (TFE). In this paper, we extend this work to include the blood component and also investigate the TFE in dynamic imaging. CT imaging and PET kinetic modelling are used to determine fractional air and blood voxel volumes in six patients with idiopathic pulmonary fibrosis. These values are used to illustrate best and worst case scenarios when interpreting images without correcting for the TFE. In addition, the fractional volumes were used to determine correction factors for the SUV and the kinetic parameters. These were then applied to the patient images. The kinetic parameters K 1 and K i along with the static parameter SUV were all found to be affected by the TFE with both air and blood providing a significant contribution to the errors. Without corrections, errors range from 34–80% in the best case and 29–96% in the worst case. In the patient data, without correcting for the TFE, regions of high density (fibrosis) appeared to have a higher uptake than lower density (normal appearing tissue), however this was reversed after air and blood correction. The proposed correction methods are vital for quantitative and relative accuracy. Without these corrections, images may be misinterpreted.