Data-Driven Control and Estimation of Medical Imaging Systems

Current medical developments depend on high quality medical imaging. To guarantee an optimal image quality, it is crucial to know the systems trajectory with high accuracy during the imaging procedure. Estimation and high performance control lead to guaranteed image qualities.

PhD Candidate: Rick van der Maas
Supervisor: Bram de Jager
Promotor: Maarten Steinbuch
Project Financing: -
Project Period: March 2012 - March 2016

Obtaining high quality 3D reconstructions of the interior of the human body, using interventional X-ray systems, is of increasing importance to the medical society. Accurate knowledge on the movement trajectories of the system is key for a successful 3D reconstruction. New developments in C-arm based X-ray systems allow for faster movements and higher reconstructions qualities. While increasing the demands on the system, inherent flexible effects typically lead to a degradation of the reconstruction quality.

In this research two approaches are investigated to guarantee an optimal reconstruction quality. High performance motion control is used to force the system to follow a predefined reference trajectory. A second approach is based on model- and observer-based estimations of the systems deviations from the ideal trajectory. By accurate modelling, system identification and control in combination with advanced imaging techniques, the complexity and required time during operation and calibration of the system is significantly reduced.