MIRD Pamphlet No. 34, Part 1: MIRDct-A Customizable Software Tool for CT Dosimetry.

Abstract

X-ray CT has long played a critical role in diagnostic radiology and is also a fundamental component of hybrid imaging in nuclear medicine, providing essential information for anatomic localization of radiopharmaceutical uptake and attenuation correction for PET/CT and SPECT/CT studies. Notably, the radiation dose to the patient from CT can be comparable to or exceed that of the administered radiopharmaceutical, underscoring the need for accurate CT dosimetry to optimize imaging protocols. With the increasing use of CT worldwide in both diagnostic radiology and nuclear medicine, robust tools are needed to assess patient radiation dose and to satisfy regulatory and human-subject research requirements. To address these needs, we developed MIRDct, a software tool integrated within the MIRDsoft.org radiation dosimetry community platform. The code offers organ model-based CT dosimetry and radiation risk metrics tailored for radiology and nuclear medicine applications. MIRDct relies on user-defined inputs for patient characteristics, scanner specifications, and technique parameters to provide rapid and accurate organ-level dose estimates. The software features built-in scan protocols with scan ranges based on predefined anatomic landmarks and supports user-defined custom scan ranges through an intuitive graphic interface. The code leverages a library of 24 state-of-the-art mesh-type phantoms modeling representative adult and pediatric patients of both sexes, configured in either arms-up or arms-down positioning (2 sexes, 6 ages, 2 arm positions). A precomputed Monte Carlo-based dose-coefficient database derived for this library enables instantaneous computation of organ-absorbed doses, effective dose, dose-length product, and size-specific dose estimates. MIRDct also provides risk-related metrics, including detriment-weighted dose, lifetime attributable risk, and risk index, and incorporates optional uncertainty quantification and error propagation to provide a quantitative indicator of estimate reliability. Dose estimates from MIRDct were benchmarked and compared across 4 scan protocols with other available codes (the National Cancer Institute dosimetry system for computed tomography and VirtualDose), demonstrating reasonable agreement (4%-30% differences for effective dose). Version 1.1, available at MIRDsoft.org, presents all underlying calculations through a transparent spreadsheet-based implementation and offers advanced features to support CT protocol optimization, deepen dosimetry insight for patient-care-related evaluations, and aid in quality assurance.