Cone-Beam computed tomography (CBCT) imaging provides better spatial resolution and exposes the patient to lower radiation doses than conventional CT-scanning, but is also subject to spacially-dependent bias due to the beam energy spectrum. This results in a very limited capacity for quantitative imaging. Routine reconstruction softwares use algorithms based on filtered backprojection, relying on rudimentary corrections with respect to the physics of the problem. These properties have a direct impact on contrast resolution and soft tissue visualization.
The goal of this project is to improve contrast resolution and reduce de presence of artéfacts on CBCT images. In order to do this, we implement a polychromatic iterative method, which allows the integration of a priori information of the system as well as the avertion of the artifacts caused by the beam hardening phenomenon and monoenergetic approximations at reconstruction. We therefore take into account the X-ray spectrum emitted by the source, the energy-dependent attenuation coefficients of selected tissues and materials, and the geometry of a cone-beam system.
Our method is based on an adaptation of IMPACT (Iterative Maximum-likelihood Polychromatic Algorithm for CT) which defines the attenuation coefficient as a linear combination of photoelectric effect and Compton effect, pondered in order to consider photon energy an the type of tissue. The implemented algorithme is tested on both synthetic and experimental phantoms, as well a clinical data.
The results show a reduction of beam-hardening artifacts such as cupping, an improved contrast in soft tissues and successful quantitative reconstruction. In clinics, this project will lead to better soft tissue imaging and therefore to greater diagnostic efficiency and accuracy.
| Date | 16 Apr 2019 |
|---|
| Original language | French |
|---|
| Awarding Institution | - École de technologie supérieure
|
|---|
| Supervisor | Jacques A. de Guise (Supervisor) & Matthieu Schmittbuhl (Co-supervisor) |
|---|
Castonguay-Henri, A. (Author),
De Guise, J. (Supervisor) & Schmittbuhl, M. (Co-supervisor),
16 Apr 2019Student thesis: Master's thesis › Master in Engineering: Engineering