3D reconstruction of cerebral blood flow and vessel morphology from x-ray rotational angiography.
Doctoral thesis, UCL (University College London).
Three-dimensional (3D) information on blood flow and vessel morphology is important when assessing cerebrovascular disease and when monitoring interventions. Rotational angiography is nowadays routinely used to determine the geometry of the cerebral vasculature. To this end, contrast agent is injected into one of the supplying arteries and the x-ray system rotates around the head of the patient while it acquires a sequence of x-ray images. Besides information on the 3D geometry, this sequence also contains information on blood flow, as it is possible to observe how the contrast agent is transported by the blood. The main goal of this thesis is to exploit this information for the quantitative analysis of blood flow. I propose a model-based method, called flow map fitting, which determines the blood flow waveform and the mean volumetric flow rate in the large cerebral arteries. The method uses a model of contrast agent transport to determine the flow parameters from the spatio-temporal progression of the contrast agent concentration, represented by a flow map. Furthermore, it overcomes artefacts due to the rotation (overlapping vessels and foreshortened vessels at some projection angles) of the c-arm using a reliability map. For the flow quantification, small changes to the clinical protocol of rotational angiography are desirable. These, however, hamper the standard 3D reconstruction. Therefore, a new method for the 3D reconstruction of the vessel morphology which is tailored to this application is also presented. To the best of my knowledge, I have presented the first quantitative results for blood flow quantification from rotational angiography. Additionally, the model-based approach overcomes several problems which are known from flow quantification methods for planar angiography. The method was mainly validated on images from different phantom experiments. In most cases, the relative error was between 5% and 10% for the volumetric mean flow rate and between 10% and 15% for the blood flow waveform. Additionally, the applicability of the flow model was shown on clinical images from planar angiographic acquisitions. From this, I conclude that the method has the potential to give quantitative estimates of blood flow parameters during cerebrovascular interventions.
|Title:||3D reconstruction of cerebral blood flow and vessel morphology from x-ray rotational angiography|
|Open access status:||An open access version is available from UCL Discovery|
|Keywords:||Blood flow quantification, rotational angiography, cerebral blood flow, 3D vessel reconstruction|
|UCL classification:||UCL > School of BEAMS > Faculty of Engineering Science > Medical Physics and Bioengineering|
Archive Staff Only