Skip to main content

Training on computational dosimetry for monitoring of Nuclear Medicine departments

Nuclear medicine personnel are often subject to high radiation doses. Some emerging evidence suggests doses, especially in case of accidental contaminations, might be higher than the dose limits for extremities (skin hand dose) or close to dose limits (eye lens). Considering the emerging role of nuclear medicine in detection and treatment of diseases and the increasing number of radioisotopes, improving the monitoring accuracy of extremities and eye lens in nuclear medicine staff is a priority in clinical practice. However, current dosemeters cannot deliver such accurate dose estimates as their sensitivity can vary greatly with different types of radiation, and they strongly depend on angular and energy distributions. Furthermore, in real clinical practice only a limited number of devices can be worn by personnel, or they may hamper their operations.

Computational dosimetry can solve these issues and deliver more accurate dose estimates. By leveraging on the capabilities of computer vision, flexible Monte Carlo simulation models and machine learning, we can calculate individualized doses for all body parts, possibly even in real time.


The objective of this training is to show how computer vision, machine learning and Monte Carlo simulations with flexible computational human model can be put together to create an online dosimetry system for monitoring nuclear medicine personnel. The training will be based on the experience gained during the development of the online dosimetry monitoring system performed in the PhD of Daniel Santiago Rondon.

On the one hand, the trainee will be presented with insights and technical information including state of the art human and object tracking, image filtering algorithms and machine learning methods to enhance the detection accuracy of complex objects. Furthermore, we will present the methodology to convert tracking data into actual simulations files (for the Gate particle transport code) including flexible computational phantoms. Attention will also be given to the techniques that might allow detection of contaminations and spills, so that the resulting exposure can be included and generate a more comprehensive dosimetry assessment in case of accident.

The minimum diploma level of the candidate needs to be

  • MD

The candidate needs to have a background in

  • Physics

Estimated duration

2 weeks

Expert group

Research in Dosimetric Applications

SCK CEN Mentor

Lombardo Pasquale
plombard [at]
+32 (0)14 33 28 63

SCK CEN Co-mentor

Santiago Rondón Daniel
dsrondon [at]
+32 (0)14