Quantitative Reality: A Phantom Tool to Assess Heterogeneity Beyond Visual Limits in Positron Emission Tomography
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background
Accurate quantification of intratumoral heterogeneity is essential for validating radiomics features and AI models in oncology. However, standard physical phantoms typically utilize homogeneous compartments, lacking the spatially complex uptake patterns required to rigorously benchmark these metrics against a known reality. To address this limitation, we introduce porous 3D-printed inserts as a solution for simulating controlled spatial heterogeneity.
Methods
In this study, we developed modular grid inserts (17, 22, 28 mm) with varying structural density to precisely modulate local radiotracer volume fraction. Nine distinct heterogeneity models were designed simulating pathological progressions from uniform uptake to necrotic cores and multifocal morphologies. The phantom was imaged on a PET/CT system using at varied activity levels for 2 different tracers: [ 18 F]FDG ([ 18 F]-2-fluoro-2-deoxy-d-glucose) and [ 68 Ga] (68-Gallium). Quantitative performance was assessed using conventional PET metrics such as Standardized Uptake Values (SUV), Recovery Coefficients (RC), and Target-to-Background Ratios (TBR).
Results
The phantom generated reproducible heterogeneous uptake patterns independent of the radiotracer used. While Partial Volume Effects (PVE) reduced visual distinctness in smaller geometries (17 mm), quantitative metrics successfully preserved the intended pathological trends. Larger inserts (28 mm) demonstrated agreement with the theoretical design. Crucially, voxel-level kernel density estimation confirmed the generation of complex, multi-peaked activity distributions necessary for testing texture analysis algorithms.
Conclusion
This study establishes a foundational ground truth for heterogeneous uptake in PET imaging. Such defined physical standard enables the robust validation of segmentation algorithms and radiomics features, ensuring quantitative reliability even in scenarios where spatial resolution limits visual assessment.
