Development of a contamination detection system in Nuclear Medicine

Introduction: Occupational radiation exposure remains a significant concern in nuclear medicine, particularly for staff involved in handling radiopharmaceuticals. Extremity dosimetry, which assesses radiation exposure to hands and fingers, is critical for ensuring the safety of healthcare workers. However, the present passive ring dosimetry systems are not sufficient to detect the maximum dose levels on the hands of the operators. Also, contaminations can occur, which risk to stay undetected.

Research at KUL and SCK CEN is on-going to develop new computational methodologies to provide rapid, individualised dose estimates for staff in NM, with special attention to extremity and skin dosimetry. This system makes use of 3D cameras and microcomputers for staff/object recognition and tracking. Flexible human models will be included to dynamically represent the staff and their posture. The digital twins will also include the radioactive source, i.e., radiopharmaceuticals within syringes and vials, with all relevant characteristics (shielding, types of emitted particles, energy spectra, activity). The expected outcomes will contribute to staff radiation protection by providing them direct feedback on their exposure, supporting the implementation of the ALARA principle and by facilitating the follow-up of the occupational dose limits. However, these computational methods do not solve the problem of detecting contaminations.

In NM, small spills of radiopharmaceuticals can lead to considerably high doses. In case the spill is contaminating the gloves or other body parts of an operator, the total radiation exposure can easily increase ten folds. Due to the chemical nature and to the high activity concentrations, minor spills are very difficult to detect, and typically they remain unnoticed until the staff member performs a contamination check with a monitor. To address this issue, a specialised system for detecting spills in the environment and contamination events will be developed. The primary objective of this system is to provide a quantitative estimation of the additional dose that such incidents could cause. A secondary objective is to alert NM staff as soon as spills or contaminations are detected.

Objective: This master thesis aims to investigate how a detection system for contaminations can be set up in nuclear medicine departments. For this active detectors will need to be installed, and the results of these detectors need to be processed in real time to detect anomalities, which would lead to the conclusion of an accidental contamination spill..

Proposed Methodology:

The detection of spills and contamination, and estimating the resulting extremity exposure, will rely on input data from environmental radiation sensors integrated into the computational framework. A multi-parameter approach based on AI will be implemented to identify anomalies in the sensor readings, which could indicate spills or contamination events. By simulating the presence of an additional incidental source (e.g., liquid spill) in the digital twins developed, the requirements relative to the performance characteristics (for example, sensitivity or energy response) of the sensors will be evaluated and their positioning in the room environment optimised. The active environmental radiation sensors will be selected based on these performance characteristics and availability on the market. The selected sensors will be tested under controlled laboratory conditions to quantify their capability for real-time monitoring of radiation source distribution and intensity.

  1. Evaluation of Detector Technologies: Conduct a comprehensive review and comparative analysis of available dosimeter technologies suitable for contamination detection
  2. Development of Calibration Protocols: Establish calibration protocols specific to the radiation energies and sources commonly encountered in nuclear medicine facilities.
  3. Field Testing and Validation Studies: Tests will be done in standard set-ups (SCK CEN laboratories) and in nuclear medicine departments (KUL) to conduct field testing of the developed detection techniques.

Expected Outcomes:

  • Identification of optimal detector technologies and calibration protocols for contamination detection in nuclear medicine.
  • Following successful lab tests and calibration, the sensors will be integrated into the dosimetry framework to feed in real-time readouts to the newly developed anomaly detection AI model.