Skip to main content
To KTH's start page To KTH's start page

Spectral Photon-Counting Computed Tomography with Silicon Detectors: New Models and Applications

Time: Fri 2022-10-21 10.15

Location: FD5, Roslagstullsbacken 21, Stockholm

Language: English

Subject area: Physics

Doctoral student: Fredrik Grönberg , Medicinsk bildfysik

Opponent: Assistant Professor Amir Pourmorteza, Emory University, USA

Supervisor: Professor Mats Danielsson, Medicinsk bildfysik; Assistant Professor Mats Persson, Medicinsk bildfysik

Export to calendar



X-ray computed tomography (CT) is a widely used imaging modality that enables visualization of nearly every part of the human body. It is used for diagnosis of disease and injury as well as medical treatment planning. The vast majority of CT scanners in clinical use today have energy-integrating x-ray detectors, which measure the total incident energy in a given measurement.Spectral photon-counting detectors operate by counting individual photons and measuring their energy, and are expected to yield the next major advance in CT, with improvements in spatial resolution, dose efficiency, material differentiation and quantitative imaging capabilities compared to the current state-of-the-art.

In this Thesis, a set of new models and applications for a spectral photon-counting silicon detector developed for CT is investigated. The first part of the Thesis is dedicated to the modeling of spectral photon-counting silicon detectors. A new statistical model for the effects of pulse pileup is presented. Also, the effects on image quality from intra-detector Compton scatter in silicon detectors are investigated via spatio-energetic modeling. In the second part of the Thesis, potential applications for spectral photon-counting detectors are investigated. An experimental study of ex vivo CT imaging of an excised human heart with calcified plaque is presented. It demonstrates the feasibility of unconstrained projection-based three-material decomposition with iodine as a third basis material and explores the potential improvements in spatial resolution and material differentiation that can be achieved with a spectral photon-counting silicon detector compared to a conventional dual-energy CTsystem. Two other applications are investigated with simulations: a method for reconstructing CT images from spectral photon-counting CT data that accurately mimic conventional CT images; and a method for estimating iron concentration in mixtures of liver and adipose tissue when using three basis functions instead of only two to describe the linear attenuation coefficient of tissues in the human body. 

Although the methods presented in this Thesis have been specifically developed for a spectral photon-counting silicon detector, they are also applicable for other types of photon-counting detectors.