Tracking of Ceiling‑Suspended Shields in Interventional Radiology and Cardiology
Introduction
Interventional radiology (IR) and interventional cardiology (IC) procedures have expanded rapidly worldwide due to their effectiveness as minimally invasive treatment techniques. However, medical staff working in these environments are exposed to relatively high levels of ionizing radiation. Occupational doses in IR and IC are among the highest in medical practice, primarily due to prolonged procedure durations, complex and highly non-uniform radiation fields, and the close proximity to both the patient and the radiation source.
Personal dosimetry is a fundamental component of radiation protection for medical staff. In current clinical practice, staff are required to wear one or multiple personal dosimeters to monitor occupational exposure.
Computational dosimetry is an emerging approach that addresses these challenges by integrating camera-based staff tracking, radiation field characterization, workplace geometry, and Monte Carlo radiation transport simulations or simplified dose calculation models. This approach has the potential to overcome the limitations associated with conventional personal dosimeters.
Ceiling-suspended protective shields play a crucial role in radiation protection in IR and IC suites. These transparent lead-acrylic shields are positioned between the patient (the primary source of scattered radiation) and medical staff in order to reduce occupational exposure. However, accurate detection and tracking of these thin, transparent objects is technically challenging. Their low visual contrast, translucency, and frequent occlusions make them difficult to detect using conventional techniques. Moreover, their protective effectiveness strongly depends on their position and orientation, both of which can change continuously during clinical procedures as staff move and equipment is repositioned. Therefore, accurate, robust, and real-time data of shield positioning is essential for reliable dose estimations.
Aim of the Thesis
The aim of this master’s thesis is to contribute to the development of a computational dosimetry system capable of estimating radiation doses to medical staff in real time, without the use of wearable dosimeters. Specifically, the thesis focuses on the accurate detection and tracking of ceiling-suspended protective shield, as their position has a significant influence on occupational radiation dose.
Project Objectives
The primary objective of this thesis is to develop and validate a computer vision-based system for tracking ceiling-suspended protective shields in IR and IC rooms. The specific objectives are:
- Shield Recognition: Develop algorithms to accurately detect and identify ceiling-suspended shields in camera images/videos under clinically realistic conditions, including varying lighting and viewing angles.
- Position and Orientation Tracking: Implement methods to determine the 3D position and orientation of the shield relative to a fixed coordinate system.
- Validation in Realistic Setups: Evaluate the performance of the tracking system in multiple realistic experimental configurations that mimic clinical scenarios, including variations in shield placements, staff positions, and equipment configurations.
Methodology
Computer Vision for Shield Detection and Tracking
The student will apply modern computer vision techniques, including deep learning–based approaches, to detect and track ceiling-suspended shields in image/video data acquired in IR/IC environments. The objective is to extract accurate shield position and orientation parameters required as input for dose calculations.
Integration of Shield Position and Orientation into the Dosimetry System
The extracted shield geometry and positioning will be translated into formats compatible with the existing dose‑calculation pipeline. This includes generating appropriate input files for Monte Carlo (MC) simulations within the computational dosimetry system. Particular emphasis will be placed on ensuring that shield position and orientation are represented accurately, such that dose calculations reflect realistic procedural conditions.
Experimental Validation
The position‑extraction methods will be validated in controlled laboratory environments that replicate typical IR/IC room geometries. Test scenarios will include variations in shield placement, orientation, and lighting conditions, as well as realistic occlusions caused by staff and equipment. Validation will focus on assessing the robustness and accuracy of the extracted shield parameters.
Expected Outcomes
Upon completion of this thesis, the student will have developed a validated prototype system for tracking ceiling-suspended protective shields, suitable for integration into a computational dosimetry framework for occupational dose assessment in IR and IC suites.