Radiation dose and x-ray beam modelling in diagnostic and interventional radiology using Monte Carlo methods

Abstract

Although medical imaging is essential for the diagnosis and treatment of a wide range of medical conditions, the radiation dose from x-ray examinations is one of the largest contributors to the exposure of the world’s population. It is therefore the responsibility of healthcare professionals to ensure that examinations are justified (a net benefit) and optimized (as low dose as reasonably achievable). In order to better understand the risks associated with medical exposures, methods for systematic and accurate patient dose estimation are necessary. This is addressed in the present thesis by introducing improved methods for x-ray beam and radiation dose modelling in diagnostic and interventional radiology. The thesis is comprised of two interconnected parts, summarized as follows.

The first part of the thesis describes the development of a deterministic model for the energy and angular distribution of x rays emitted from an x-ray tube. The model combines Monte Carlo-calculated results and theoretical physics data to account for the depth, energy, and angular distribution of bremsstrahlung and characteristic x rays produced in an x-ray tube anode. The model is an improvement over previous models, especially for low kilovoltage x-ray beams (below 50 kV), and it is reliable for a broader angular distribution of the x-ray emission, making it suitable for the prediction of central-axis spectra, as well as off-axis effects such as the (anode) heel effect. It is able to reproduce narrow-beam Monte Carlo calculations to within 0.5% in terms of the aluminum half-value layer thickness (HVL), and is in good agreement (< 2% in HVL) with measured spectra for typical diagnostic and therapeutic x-ray beams.

The second part concerns the development and application of a framework for systematic estimation of patient organ absorbed doses. The framework includes a method for reconstructing the exposure geometry based on non-proprietary access to widely available radiation dose structured reports (RDSR). By combining the framework with an x-ray source model (such as the one developed in this work), and Monte Carlo simulations of radiation transport, systematic estimation of patient doses is possible, something that has traditionally been difficult to achieve. A prototype implementation of the framework is demonstrated for selected radiation dose estimations. The applications are shown to provide an enhanced understanding of the patient radiation exposure and the risks associated with radiation, which is useful for the optimization of clinical methods and protocols.

The methods developed in this work can be used by healthcare professionals as well as researchers to justify and optimize each medical x-ray exposure performed in the treatment of a patient, thereby ensuring the safe use of radiation in medicine.