Improvements in brain imaging using spectral computed tomography: Development and evaluation of novel techniques

Background and aimStroke is a life-threatening condition requiring timely and accurate treatment. Computed tomography (CT) is widely used to detect stroke due to its accessibility, speed and diagnostic value. Spectral CT, the latest advancement in CT technology, utilizes multiple energy levels to improve tissue contrast and material differentiation. This thesis aims to evaluate how spectral reconstructions – monoenergetic images (MI) and iodine quantification – can improve brain imaging for stroke assessment.MethodsPaper I assessed the feasibility of multiphase CT angiography (mCTA) for quantitative brain perfusion analysis, using CT perfusion (CTP) as the reference standard. Paper II compared the diagnostic performance of MIs and conventional polyenergetic images (PI) for detecting ischemia using non-contrast CT (NCCT). Paper III assessed whether contrast media (CM) dose during CTA could be reduced by 50% while maintaining the same image quality as using PIs with a standard CM dose. Paper IV evaluated the effectiveness of using metal artifact reduction (MAR) and MIs to reduce metal artifacts from intracranial coils. All studies were retrospective and, except for Paper I, included both quantitative and qualitative analyzis.ResultsPaper I: mCTA demonstrated potential for perfusion deficit detection, with MIs at 40 keV and iodine quantification providing better separation between normal and underperfused tissue than PIs. Paper II: Mis reconstructed at 50–60 keV improved ischemia detection and diagnostic accuracy compared to PIs. Paper III: MIs reconstructed at 50 keV can be used to preserve image quality while reducing CM dose by 50%. With CM dose at 100%, using MIs improved image quality compared to PIs and can be used to salvage scans with poor contrast enhancement. Paper IV: High-energy MIs (≥120 keV) with MAR reduced metal artifacts, though lower-energy MIs (60–100 keV) were superior for soft tissue contrast. An intermediate level of 80 keV could provide an optimal balance.ConclusionsSpectral CT improves stroke imaging by enhancing perfusion assessment, soft tissue contrast, ischemia detection and vascular visualization, and reducing metal artifacts