Iterative reconstruction for mobile chest tomosynthesis

Date: 17 January 2017

Venue: UAntwerp, Campus Drie Eiken, Promotiezaal Q0.02 - Universiteitsplein 1 - 2610 Antwerpen-Wilrijk (route: UAntwerpen, Campus Drie Eiken)

Time: 5:00 PM

Organization / co-organization: Department of Physics

Principal investigator: Jan Sijbers & Gert Behiels

Short description: PhD defence Jeroen Cant - Faculty of Science - Department of Physics


The chest X-ray exam (CXR) is one of the oldest medical imaging techniques. Thanks to the low radiation dose and relatively cheap cost compared to CT and other techniques, it is still frequently used to get a first insight into the condition of the patient. For the monitoring of less mobile patients, the CXR is performed bedside with a portable detector and a mobile X-ray unit.

However, the interpretation of the CXR is often complicated by overlapping anatomy and suboptimal positioning, especially in bedside exams. Chest tomosynthesis (CTS) solves this problem of anatomical noise. CTS is an X-ray imaging technique in which a small number of low dose chest X-ray images are acquired from different angles, using a motorized X-ray tube. Subsequently, section images are computed parallel to the detector by a reconstruction algorithm. The overlapping anatomy is thereby separated into different section images, which improves the detection of lung nodules and simplifies the detection of various other abnormalities in the chest, compared to a CXR.

CTS is currently only available on a fixed modality in the radiology department. In this work, several challenges are dealt with to enable the development of a mobile version of CTS, based on a portable detector and a mobile X-ray unit. Techniques were developed for reconstruction with a reduced number of projection images, and for the automatic calibration of the mobile geometry. Finally, the clinical potential of mobile CTS was illustrated in a simulation study for patients in an Intensive Care Unit.