Uppskattning av mjukvävnadselasticitet i CT-bilder

In the United states, liver disease is the 10th leading cause of death and affects 4.5 million adults, resulting in 16 deaths per 100,000 people annually [1]. Globally, Chronic Liver Disease (CLD) impacts more than 800 million people and results in approximately two million deaths per year [2]. Liver stiffness, a characteristic of CLD, is able to be measured via elastography [3]. Elastography methods are well established in Ultrasound (US) and Magnetic Resonance Imaging (MRI) modalities, but remains underexplored in Computed Tomography (CT). This project aimed at developing software capable of providing semi-quantitative analysis of regional tissue heterogeneity using multi-phase CTscans. To achieve this goal, various registration algorithms were evaluated, methods to improve registration performance were investigated, and the correlation between deformation vector fields and soft tissue elasticity was assessed. B-spline based methods with vessel region exclusion demonstrated the best performance when aligning the pre-contrast phase to the arterial phase and the portal-venous phase to the arterial phase. Registration with vessel region exclusion displayed potential to improve alignment between phases, but requires additional investigation. Further, the ability of deformation vector f ields attained from registration to discern cirrhotic and non-cirrhotic groups showed promise, but needs to be tested on a larger cohort to validate results. Significant differences between these groups were seen for mean and median surface distance when registering the portal-venous phase to the pre-contrast phase using a Diffeomorphic Demons algorithm. A novel approach to estimating relative tissue stiffness in the liver using multi-phase CT was introduced and shows potential to expand the scope of elastography modalities beyond US, and MRI. An effective CT based elastography method could increase care accessibility to patients who cannot undergo other modalities or for whom these modalities are less effective. CT based elastography has the potential to improve early detection and diagnosis of liver cirrhosis, benefiting millions of patients worldwide.