Are quantitative features of lung nodules reproducible at different CT acquisition and reconstruction parameters?

Abstract

Consistency and duplicability in Computed Tomography (CT) output is essential to quantitative imaging for lung cancer detection and monitoring. This study of CT-detected lung nodules investigated the reproducibility of volume-, density-, and texture-based features (outcome variables) over routine ranges of radiation dose, reconstruction kernel, and slice thickness. CT raw data of 23 nodules were reconstructed using 320 acquisition/reconstruction conditions (combinations of 4 doses, 10 kernels, and 8 thicknesses). Scans at 12.5%, 25%, and 50% of protocol dose were simulated; reduced-dose and full-dose data were reconstructed using conventional filtered back-projection and iterative-reconstruction kernels at a range of thicknesses (0.6–5.0 mm). Full-dose/B50f kernel reconstructions underwent expert segmentation for reference Region-Of-Interest (ROI) and nodule volume per thickness; each ROI was applied to 40 corresponding images (combinations of 4 doses and 10 kernels). Typical texture analysis metrics (including 5 histogram features, 13 Gray Level Co-occurrence Matrix, 5 Run Length Matrix, 2 Neighboring Gray-Level Dependence Matrix, and 3 Neighborhood Gray-Tone Difference Matrix) were computed per ROI. Reconstruction conditions resulting in no significant change in volume, density, or texture metrics were identified as “compatible pairs” for a given outcome variable. Our results indicate that as thickness increases, volumetric reproducibility decreases, while reproducibility of histogram- and texture-based features across different acquisition and reconstruction parameters improves. To achieve concomitant reproducibility of volumetric and radiomic results across studies, balanced standardization of the imaging acquisition parameters is required.

A Non-invasive Method to Diagnose Lung Adenocarcinoma

“…We used stratified random sampling to balance the covariates. In feature selection, A few issues regarding the stability and reproducibility of the radiomics features have been raised in recent years (31)(32)(33). Multiple parameter changes (e.g., slice thickness) in general produce greater measurement errors.…”

A systematic review and quality of reporting checklist for repeatability and reproducibility of radiomic features

“…To mitigate this effect, Park et al demonstrated that CT feature reproducibility can be improved by using a CNN-based super resolution algorithm [31] . For CT studies, Erdal et al pointed out that a slice thickness of 2 mm leads to the most accurate shape features [32] . It remains unclear if radiomic features extracted from images acquired under different conditions also lead to different conclusions.…”

“…Also in CT studies, contours delineated by different clinicians impacted the prognostic value of radiomic features. For head and neck cancer, in one PET and one MR study features were found to be very sensitive to differences in tumor delineation [32] , [33] , [37] , [44] . Overall, semi-automated and automated algorithms produced more stable results.…”

Influence of Adaptive Statistical Iterative Reconstructions on CT Radiomic Features in Oncologic Patients

“…An important aspect is the CTTA reproducibility related to the influence of CT acquisition parameters (e.g., level of radiation dose, slice thickness, reconstruction algorithms) that can affect results and standardization among different Diagnostics 2021, 11, 1000 2 of 10 studies [8][9][10]. This aspect is gaining interest, as shown by Erdal et al [11] who demonstrated that slice thickness influenced reproducibility of radiomic features in lung nodules, and Prezzi et al [12] who showed the influence of iterative reconstruction (IR) algorithm versus traditional filtered back projection (FBP) on radiomics quantification in twenty-eight datasets of colorectal cancer.…”