Application of Partial Volume Effect Correction and 4D PET in the Quantification of FDG Avid Lung Lesions

• Published in: Molecular imaging and biology Vol. 17; no. 1; pp. 140 - 148
• Main Authors: Salavati, Ali, Borofsky, Samuel, Boon-Keng, Teo K., Houshmand, Sina, Khiewvan, Benjapa, Saboury, Babak, Codreanu, Ion, Torigian, Drew A., Zaidi, Habib, Alavi, Abass
• Format: Journal Article
• Language: English
• Published: Boston Springer US 01-02-2015; Springer Nature B.V
• Subjects: Aged; Algorithms; Automation; Contrast Media; Female; Fluorodeoxyglucose F18; Humans; Image Processing, Computer-Assisted; Imaging; Lung - diagnostic imaging; Lung - pathology; Lung Neoplasms - diagnostic imaging; Lung Neoplasms - radiotherapy; Male; Medicine; Medicine & Public Health; Middle Aged; Motion; Multimodal Imaging; Multivariate Analysis; Positron-Emission Tomography; Radiology; Radiopharmaceuticals; Regression Analysis; Reproducibility of Results; Research Article; Respiration; Tomography, X-Ray Computed; Whole Body Imaging; FDG-PET/CT; Respiratory gating; Metabolic activity quantification; Partial volume effect; Lung cancer
• ISSN: 1536-1632; 1860-2002
• DOI: 10.1007/s11307-014-0776-6

Purpose
The aim of this study is to assess a software-based method with semiautomated correction for partial volume effect (PVE) to quantify the metabolic activity of pulmonary malignancies in patients who underwent non-gated and respiratory-gated 2-deoxy-2-[
18
F]fluoro-
d
-glucose (FDG)-positron emission tomography (PET)/x-ray computed tomography(CT).
Procedures
The study included 106 lesions of 55 lung cancer patients who underwent respiratory-gated FDG-PET/CT for radiation therapy treatment planning. Volumetric PET/CT parameters were determined by using 4D PET/CT and non-gated PET/CT images. We used a semiautomated program employing an adaptive contrast-oriented thresholding algorithm for lesion delineation as well as a lesion-based partial volume effect correction algorithm. We compared respiratory-gated parameters with non-gated parameters by using pairwise comparison and interclass correlation coefficient assessment. In a multivariable regression analysis, we also examined factors, which can affect quantification accuracy, including the size of lesion and the location of tumor.
Results
This study showed that quantification of volumetric parameters of 4D PET/CT images using an adaptive contrast-oriented thresholding algorithm and 3D lesion-based partial volume correction is feasible. We observed slight increase in FDG uptake by using PET/CT volumetric parameters in comparison of highest respiratory-gated values with non-gated values. After correction for partial volume effect, the mean standardized uptake value (SUVmean) and total lesion glycolysis (TLG) increased substantially (
p
value <0.001). However, we did not observe a clinically significant difference between partial volume corrected parameters of respiratory-gated and non-gated PET/CT scans. Regression analysis showed that tumor volume was the main predictor of quantification inaccuracy caused by partial volume effect.
Conclusions
Based on this study, assessment of volumetric PET/CT parameters and partial volume effect correction for accurate quantification of lung malignant lesions by using respiratory non-gated PET images are feasible and it is comparable to gated measurements. Partial volume correction increased both the respiratory-gated and non-gated values significantly and appears to be the dominant source of quantification error of lung lesions.