Quantitative imaging of vulnerable atherosclerotic plaque using MARS Spectral CT

This thesis details the novel use of the Medipix All Resolution System (MARS) Spectral CT scanner for quantitative imaging of vulnerable atherosclerotic plaque. The motivation behind this research was to see how well the MARS scanner could identify and quantify common com- ponents that are present in atheromas. Characteristic features of vulnerable plaque include a large necrotic lipid core, thin fibrous cap, increased inflammation, and intraplaque haemorrhage. The ability to identify lipid was investigated by joining high-Z contrast agents to low-density lipoproteins, enhancing the signal of a lipid component using spectral CT. Contrast agents were also implemented using U937 cells and macrophages, following the progressive lifecycle of the disease. The elements iron and calcium were investigated in relation to their similar spectral pro- files. With microcalcifications attributed to plaque instability, and iron a key component in blood, and hence intraplaque haemorrhage, the ability to distinguish these materials is crucial for identifying vulnerable plaque features in a clinical setting. Phantom studies involving iron- and calcium- like materials were performed to optimise the phantom and scanning protocol. Misidentifications between specific material concentrations in the material decomposition un- derwent further analysis with new methods that used the energy information from the scan. This revealed that misidentified materials could be separated and distinguished with the spec- tral information, demonstrating the advantage of spectral CT for atherosclerosis imaging over other imaging modalities. With the coexistence of iron and calcium deposits in atherosclerotic plaque, a point of discussion surrounds the issue of whether the MARS scanner can identify these two materials simultaneously, given there is a large enough signal. This was investigated through a mixed phantom of iron- and calcium- like solutions, to determine the point at which the iron signal predominates, and predominates over every other signal. In time, improvements to the scanner and the material decomposition software will lead to reduced misidentification of materials, and stronger distinction between iron and calcium . The experiments performed throughout this thesis contribute to the field of atherosclerosis and the capabilities of the MARS Spectral CT scanner to identify and quantify elemental features of vulnerable plaque. With the MARS team fast approaching the goal of bringing multi- energy CT to human scanning, this research establishes firm ground for continued atheroma studies to help bring spectral CT to clinical imaging of atherosclerosis