Room: Track 1
Purpose: To developed a cross-platform app that automates the workflow of a radiological exposure assessment to medically exposed patients from radiographic and fluoroscopic imaging procedures in large-scale epidemiological studies.
Methods: The cross-platform app prompts the user to enter patient data such as age (1-, 5-, 10- , 15-year-old, or adult), sex (female or male), height, and weight. The technical parameters required to correctly model the medical exposure are beam quality (HVL range = 0.66–10 mm of Al; peak tube potential = 50–120 kVp; added filtration = 0.0 mm Al to 0.9 mm Cu), predefined protocol (e.g., pneumothorax, pneumoperitoneum, cardiac, or gastro-intestinal study procedure), and geometry information. The comprehensive range of values encompasses medical exposures dating from the 1930’s—accommodating established epidemiological cohorts—to contemporaneous clinical practice. The app generates input files and job submission scripts to run Monte Carlo radiation transport simulations using MCNP coupled with a hybrid phantom–selected from a library of 20 phantoms–that closely matches the patient data. Output files are then processed to report the energy deposition to 58 tissues—including the heart and associated structures, lymph nodes, and active bone marrow.
Results: Our app relies on minimal user input and draws most of the information from anatomical landmarks on the hybrid phantoms. Validity of the code was checked with assertions statements, test cases, and visual representations in ImageJ and MCNP VisEd. We used this workflow in the exposure assessment of the Canadian Fluoroscopy Cohort Study. For this assessment, we were able to generate 3,240 unique input files and run the subsequent transport calculations in under a week.
Conclusion: Our cross-platform app, although currently limited to a few predefined protocols, greatly facilitates the radiological exposure assessment in large-scale epidemiological studies.