Interobserver and Intraobserver Variability among Measurements of FDG PET/CT Parameters in Pulmonary Tumors

Abstract

Background: 18F-fluorodeoxyglucose (FDG) positron emission tomography computed tomography (PET/CT) provides information about metabolic and morphologic status of malignancies. Tumor size and standardized uptake value (SUV) measurements are crucial for cancer treatment monitoring. Aims: The purpose of our study was to assess the variability of these measurements performed by observers evaluating lung tumors. Study Design: Retrospective cross-sectional study. Methods: FDG PET/CT images of 97 patients with pulmonary tumors were independently evaluated by two experienced nuclear medicine physicians. Primary tumor size (UDCT), maximum SUV (SUVmax), mean SUV (SUVmean) and maximum SUV normalized to liver mean SUV (SUVnliv max) were measured by each observer at two different times with an interval of at least 2 weeks. Interobserver and intraobserver variabilities of measurements were evaluated through statistical methods. Results: Size of the lesions varied from 0.81 to 13.6 cm (mean 4.29±2.24 cm). Very good agreement was shown with correlation, Bland-Altman and regression analysis for all measured PET/CT parameters. In the interobserver and intraobserver variability analysis, the Pearson correlation coefficients were greater than 0.96 and 0.98, respectively. Conclusion: Semi-quantitative measurements of pulmonary tumors were highly reproducible when determined by experienced physicians with clinically available software for routine FDG PET/CT evaluation. Consistency may be improved if the same observer performs serial measurements for any one patient.

Impact of solitary pulmonary nodule size on qualitative and quantitative assessment using 18F-fluorodeoxyglucose PET/CT: the SPUTNIK trial

“…This persisted even after removing cases that may have been measured using a different reconstruction algorithm at the reporting and Core lab sites. This degree of variability is in contra-distinction to previous reports where the limits of agreement are typically ± 0.5–0.8 [ 25 , 26 ]. However, these results were for readers all within a single centre using a single scanner and single post-processing software.…”

Discovery of pre-therapy 2-deoxy-2-18F-fluoro-D-glucose positron emission tomography-based radiomics classifiers of survival outcome in non-small-cell lung cancer patients

“…Intra- and inter-observer variability of radiomics features from 18 randomly selected patients are displayed in Supplementary Tables 5 and 6 . There was near-perfect [ 15 , 16 ] intra- and inter-observer variability in the PET-derived radiomics features. The intra-observer variability of the radiomics features alone and when combined with PET features was 0.9 and 0.92 respectively.…”

Optimizing SUV Analysis: A Multicenter Study on Preclinical FDG-PET/CT Highlights the Impact of Standardization

“…For clinical PET/CT imaging, several studies have assessed inter-and intraobserver variability and proposed methods to standardize image analysis [7][8][9][10]. Until now, there hasn't been any study conducted on preclinical PET/ CT imaging that includes a standardized image analysis.…”