Purpose: In interventional radiography procedures, the radiologist receives relatively high radiation dose to the eye lens and skin on the hands. The conventional dosimetry, which is mostly based on a personal dosimeter worn under the lead apron, is inappropriate to assess the dose to eye lens or exposed skin. In the present study, a real-time computational dose calculation system was developed and tested, which calculates the lens and skin dose distribution realistically considering the continuously changing conditions of the X-ray machine, shields, and radiologist.
Methods: McDCIR system consists of three main parts: machine tracking, body tracking, and dose calculation. The machine tracking part obtains the continuously changing conditions of the X-ray machine and shields. The body tracking part obtains the motion data of the radiologist by using a depth camera and a body-tracking algorithm. The dose calculation part simulates the continuously changing conditions of the X-ray machine, shields, and radiologist and then calculates radiation dose to the radiologist in real-time using a library of pre-calculated 3D dose maps.
Results: The developed system was tested for a mockup procedure in the actual operating room using an X-ray imaging system (Allura Xper FD20, Philips, Netherlands). The GUI window successfully showed the mesh-type human phantom being deformed in real-time according to the motion of the radiologist with color-coded visualization of the skin dose rate distribution. The information panel showed detailed dose values including the lens dose and the average and maximum skin doses. The test was successful, demonstrating the real-time operation of the entire system, including real-time machine and body tracking and accurate dose calculations.
Conclusion: The system will be tested for clinical cases, by comparing the results from McDCIR system with the measured data. McDCIR is expected to provide reliable dose values for interventional radiologists.