Effect of sampling frequency on perfusion values in perfusion CT of lung tumors. AJR Am J Roentgenol 2013 Feb;200(2):W155-62
Date
01/25/2013Pubmed ID
23345379Pubmed Central ID
PMC3880201DOI
10.2214/AJR.12.8664Scopus ID
2-s2.0-84873584621 (requires institutional sign-in at Scopus site) 13 CitationsAbstract
OBJECTIVE: The purpose of this study was to assess as a potential means of limiting radiation exposure the effect on perfusion CT values of increasing sampling intervals in lung perfusion CT acquisition.
SUBJECTS AND METHODS: Lung perfusion CT datasets in patients with lung tumors (> 2.5 cm diameter) were analyzed by distributed parameter modeling to yield tumor blood flow, blood volume, mean transit time, and permeability values. Scans were obtained 2-7 days apart with a 16-MDCT scanner without intervening therapy. Linear mixed-model analyses were used to compare perfusion CT values for the reference standard sampling interval of 0.5 second with those of datasets obtained at sampling intervals of 1, 2, and 3 seconds, which included relative shifts to account for uncertainty in preenhancement set points. Scan-rescan reproducibility was assessed by between-visit coefficient of variation.
RESULTS: Twenty-four lung perfusion CT datasets in 12 patients were analyzed. With increasing sampling interval, mean and 95% CI blood flow and blood volume values were increasingly overestimated by up to 14% (95% CI, 11-18%) and 8% (95% CI, 5-11%) at the 3-second sampling interval, and mean transit time and permeability values were underestimated by up to 11% (95% CI, 9-13%) and 3% (95% CI, 1-6%) compared with the results in the standard sampling interval of 0.5 second. The differences were significant for blood flow, blood volume, and mean transit time for sampling intervals of 2 and 3 seconds (p ≤ 0.0002) but not for the 1-second sampling interval. The between-visit coefficient of variation increased with subsampling for blood flow (32.9-34.2%), blood volume (27.1-33.5%), and permeability (39.0-42.4%) compared with the values in the 0.5-second sampling interval (21.3%, 23.6%, and 32.2%).
CONCLUSION: Increasing sampling intervals beyond 1 second yields significantly different perfusion CT parameter values compared with the reference standard (up to 18% for 3 seconds of sampling). Scan-rescan reproducibility is also adversely affected.
Author List
Ng CS, Chandler AG, Wei W, Anderson EF, Herron DH, Kurzrock R, Charnsangavej CAuthor
Razelle Kurzrock MD Center Associate Director, Professor in the Medicine department at Medical College of WisconsinMESH terms used to index this publication - Major topics in bold
AdultAged
Blood Flow Velocity
Blood Volume
Female
Humans
Linear Models
Lung Neoplasms
Male
Middle Aged
Prospective Studies
Radiographic Image Interpretation, Computer-Assisted
Reproducibility of Results
Tomography, X-Ray Computed