Author: Belen Juste, Choonsik Lee, Matthew Mille, Rafael Miró, Sergio Morato Rafet, Agustin Santos, Gumersindo Verdú 👨🔬
Affiliation: Division of Cancer Epidemiology and Genetics, National Cancer Institute, Universitat Politècnica de València, Servicio de Radiofísica y Protección Radiológica, Consorcio Hospitalario Provincial de Castellón, National Cancer Institute 🌍
Purpose: To evaluate the differences in CT scan radiation dose estimation between personalized dose reconstruction, based on real patient CT images, and generalized phantom-based dose calculations.
Methods: A personalized dose reconstruction methodology was developed using anatomical models derived from real patient CT images. Validation was performed by comparing point dose measurements, acquired with MOSFET dosimeters in a Rando phantom during a Siemens Somatom Emotion CT scan, with Monte Carlo (MC) simulations. A 3D voxelized patient-specific model was created from DICOM data and used as input for MC-GPU code, with simulation parameters extracted automatically. The methodology was applied to two real cases. In the first case, an anatomical extension technique developed in previous works has been applied to realistically extend the partially captured patient anatomy when the organs are partially represented in the CT images. In the second case, a realistic case that uses a complete CT scan anatomy considering 134 anatomical structures or organs have been presented.
Results: The MC simulation results demonstrated strong agreement with the experimental dose measurements in the phantom, confirming the accuracy of the proposed methodology. The anatomy extension case reveals the overestimation of the radiation dose (averaged for the entire mass of the organ) calculated using methods that rely on patient images covering partially some organs. When compared to NCICT calculations, significant differences in organ doses were observed, ranging from 20-30% for some organs. These differences underscore the importance of personalizing dose calculations to account for individual patient anatomy.
Conclusion: This study highlights the advantages of personalized dose reconstruction over conventional phantom-based methods. The developed methodology accurately accounts for patient-specific anatomy and scan parameters, enabling more precise organ dose estimation. Personalized approaches are essential for reducing errors in dose calculations, which can have significant implications for patient organ dose estimation and future epidemiology studies.