MCNP Studies of Doses from Clinical X-ray Procedures

Kenneth A. Van Riper
White Rock Science
and
Robert L. Metzger
Radiation Safety Engineering, Inc.


Published in: Transactions of the American Nuclear Society, 75, 142 (1996)

Radiation doses from clinical x-ray examinations are a concern for the patient and for persons in and adjacent to the examination room. We have been using MCNP to calculate exposures in a variety of scenarios.

A previous work1 considered nonoccupational doses in rooms and corridors adjacent to the examination room. We considered the adequacy of existing shielding in light of proposed lowering of dose limits by a factor of five. The model incorporated all materials that would normally attenuate the beam in an x-ray room, including the patient, film canister, walls, ceiling, and floor. We included realistic assumptions for work load and a variety of tube potentials based on actual usage. Ring detectors were placed at several locations outside the room. We found that the shielding in place in most current x-ray rooms meets the new standards.

We are currently considering organ doses from diagnostic radiology during invasive procedures done under fluorsocopic guidance. These procedures can involve prolonged exposure times. High dose rates may be necessary when thick body parts, such as an oblique hip, must be penetrated. While the dose at skin entrance may be accurately estimated from measurements in air, the doses to organs within the primary beam and adjacent to the area traversed by the fluoroscope are difficult to calculate. Unlike common x-ray projections used in film radiology, the position and movement of the fluoroscopic x-ray beam are not standardized for most procedures and routine tables of organ doses cannot be used.

We are using MCNP to calculate organ doses during fluoroscopic procedures. To speed and simplify the preparation of the MCNP input file, standard patient phantom models2and energy spectra for common x-ray tube potentials3,4were assembled as 'building blocks' that can be assembled to model a specific procedure. Since the use of the standard thin phantom can overestimate organ doses by a factor of two in a heavy patient (12 cm thicker than the phantom torso), the dimensions (along with the sex and age) of the phantom are adjustable. We are developing a program to permit a user to assemble these building blocks with a simple graphical user interface.

One application of the model involves a fluoroscopic procedure in which a stent was guided through the patient's bladder and right ureter to clear an obstruction in the ureter. Because the patient was three weeks pregnant, the dose to the uterus was of interest. The exposure time and intensity and other characteristics were recorded during the procedure and were reflected in the MCNP input. Several calculations were made for different beam positions aimed between the right kidney and bladder. Visualization of particle tracks using the Sabrina code, as shown in Figure 1, verified the position of the beam. The tallies at skin entrance and in the uterus converged after 600,000 histories. Comparison with a point kernel code showed a Monte Carlo model is essential for accurate results when the organ of interest is outside the primary beam and the dose is due to scattered x-rays. We did not use any variance reduction methods for this model; such techniques would be necessary for evaluating doses to distant organs, such as the thyroid. Each calculation required several hours of running time on a Pentium-based PC.


Figures

These images have suffered horribly from multiple format conversions.

SABRINA rendering of the geometry and particle tracks for a fluoroscope model. The first 100 particle tracks are shown and the skin and undifferentiated flesh are drawn transparent to visualize the internal organs.

Upper Fluoroscope Position.

Lower Fluorscope Position.

Side View of Upper Fluorscope Position.

Model of ERCP Procedure.


These methods, while applied to diagnostic imaging special procedures, may also be used to evaluate organ doses from any external source of radiation, including mixed field radiation. The techniques may be used to determine an accurate embryo/fetal dose from external fields. Currently, dose to the embryo/fetus of a declared pregnant worker is based on film or TLD personnel monitors that commonly overestimate the actual dose.

The method can also aid in the evaluation of exposures where the personnel monitor may not be positioned to properly estimate a whole body dose, such as sources located above or below the worker, or a dose delivered from the rear of the worker where the film badge would be shielded by the employee's body.


1R. L. Metzger, R. Richardson, K. A. Van Riper, "A Monte Carlo Model for Retrospective Analysis of Shield Design in a Diagnostic XRay Room." Health Physics, 65, 164 (1993).

2K. A. Van Riper, D. L. Spikes, R. G. McDowell, "Variable Human Anthropomorphic Models," Advances and Applications in Radiation Protection and Shielding (American Nuclear Society: La Grange Park, IL), 2, 643 (1996).

3T. R. Fewell, R. E. Shuping, "The Photon Energy Distribution of some Typical Diagnostic X-Ray Beams," Medical Physics, 4, 187 (1977).

4T. R. Fewell, R. E. Shuping, "Handbook of Computed Tomography X-Ray Spectra," HHS Publication (FDA) 81-8162 (1981).