Formulation and Characterization of Gypsum and Silicone-Based Materials Based on Hounsfield Unit Values as a Basis For Developing Anthropomorphic Phantoms on Ct-Scan

Abstract

Computed tomography (CT) imaging requires phantoms with tissue-equivalent Hounsfield Unit (HU) values, but commercial anthropomorphic phantoms are expensive and have limited flexibility. This study aims to formulate and characterize gypsum and silicone-based materials based on HU values ​​as a basis for developing anthropomorphic CT phantoms. The study used a quantitative Research and Development (R&D) design with the ADDIE model. The population included potential materials for phantoms and anonymized CT images of patient heads. The sample consisted of six material formulas (two gypsum, four silicone) and 20 patient CT images. Instruments included a multislice CT scanner, a DICOM viewer with ROI tools, and precision measurement equipment. Data analysis used descriptive statistics on density, mean HU, percentage error, coefficient of variation (CV), HU difference between ROIs, and physical stability. The results showed that five of the six formulas met the accuracy criteria (error ≤15%), namely cortical gypsum (1.2%), trabecular gypsum (4.8%), muscle silicone (0.5%), gray matter silicone (9.9%), and white matter silicone (5.9%). Skin silicone exceeded the limit (26.2%). All materials showed excellent repeatability (CV 0.15–1.10%). The silicone material was homogeneous (HU difference ≤5 HU), while the gypsum was not homogeneous (12.7–13.0 HU). All phantoms were physically stable. Gypsum and silicone materials have the potential to be economical candidates for anthropomorphic CT phantoms, with the need for optimization of the skin silicone formula and gypsum homogeneity.