Preclinical X-ray imaging beyond attenuation contrast

Sammanfattning: Medical imaging is a cornerstone of modern clinical practice. Here, X-ray imaging is the given choice for rapid morphological imaging with excellent spatial resolution, albeit with sensitivity often insufficient for resolving subtle pathological changes to soft tissues. Fundamentally, the sensitivity issue is due to the image contrast traditionally being based on differential X-ray attenuation (i.e., absorption and scattering) where attenuation properties of soft tissues are often very similar. Improving the sensitivity of clinical X-ray imaging therefore requires moving beyond conventional attenuation contrast.Motivated by the above, this Thesis explores two alternative contrast mechanisms in the preclinical domain, yet with a clinical outlook: X-ray fluorescence and X-ray phase contrast. These mechanisms are demonstrated both experimentally on animal models (in vivo) and computationally on virtual anatomical phantoms (in silico). Specifically, we developed instrumentation for in vivo X-ray fluorescence imaging of mice injected with nanoparticle contrast agents, demonstrating a path towards molecular X-ray imaging with higher spatial resolution (< 0.5 mm) than established molecular modalities (e.g., PET & SPECT) and roughly 10× higher sensitivity (~ 0.1 mM) compared to conventional attenuation contrast. Furthermore, we showed that the terminal bronchioles (diameters down to ~ 60 μm) could be resolved in free-breathing mice under anesthesia using X-ray imaging boosted by phase contrast. Lastly, we showed through in silico modeling that the extension of X-ray phase contrast to human lungs could potentially enable visualization of small airways (diameters below 2 mm) which are invisible to attenuation contrast alone. In summary, this Thesis provides experimental and computational demonstrations indicating that both X-ray fluorescence and X-ray phase contrast could provide a path towards clinical X-ray imaging with improved sensitivity.