Investigation of Robustness of Radiomics Features Generated With Grey Level Co-occurrence Matrix (GLCM) for Positron Emission Tomography (PET) Image Analysis
Abstract Background: Quantification of heterogeneous radiotracer uptake in PET has the potential to be used as a biomarker of prognosis. Textural features accounting for both spatial and intensity information have recently been applied to FDG-PET images and used to predict treatment response. However, textural features have been predicted to strongly depend on volume. Other factors affecting textural features such as segmentation and quantization have previously been investigated on clinical data while image contrast and noise have not been assessed systematically. This study aims to investigate the relationships between textural features and these factors using phantom data.Methods: The torso NEMA phantom was first filled with 18F solutions to yield different contrasts between the six hot spheres (0.5-27 cm3) and the colder uniform background (2:1, 4:1, 8:1) and scanned on the TrueV PET-CT scanner for 120min. Images were reconstructed using OSEM (4 iterations, 21 subsets) for different scan durations (15-120min) and smoothed with a 4-mm Gaussian filter. The phantom with two heterogeneous spherical inserts (8.2 and 18.8 cm3) was then scanned and reconstructed using same protocol for contrast 4:1 only. All spheres were delineated using three approaches 1) the exact boundaries based on their known diameters, 2) 40% fixed threshold and 3) adaptive threshold. Textural features were derived from the co-occurrence matrix using different quantization levels (8-256). Results: Some textural features (contrast, dissimilarity, entropy, correlation) increase while others (homogeneity, energy) decrease with quantization at different rates depending on sphere volume. When using the exact delineation, contrast and scan duration (noise) have a lesser effect on textural features than sphere volume. When applying the same exact regions on the uniform background (no partial volume), the relationships between textural features and volume are comparable to when applied to the respective spheres except for correlation. Textural features are indirectly related to noise and contrast via segmentation with adaptive threshold being superior compared to the fixed threshold. Conclusion:Among the six textural features, homogeneity and dissimilarity are the most suitable for measuring PET tumour heterogeneity with quantization 64 if regions are segmented using methods that are robust to noise and contrast variations. To use these textural features as prognostic biomarkers, changes in textural features between baseline and treatment scans should always be reported along with the changes in volumes.