Herbert Kemeth

Towards cardiac C-arm computed tomography

Cardiac interventional procedures would benefit tremendously from sophisticated three-dimensional image guidance. Such procedures are typically performed with C-arm angiography systems, and tomographic imaging is currently available only by using preprocedural computed tomography (CT) or magnetic resonance imaging (MRI) scans. Recent developments in C-arm CT (Angiographic CT) allow three-dimensional (3-D) imaging of low contrast details with angiography imaging systems for noncardiac applications. We propose a new approach for cardiac imaging that takes advantage of this improved contrast resolution and is based on intravenous contrast injection. The method is an analogue to multisegment reconstruction in cardiac CT adapted to the much slower rotational speed of C-arm CT. Motion of the heart is considered in the reconstruction process by retrospective electrocardiogram (ECG)-gating, using only projections acquired at a similar heart phase. A series of N almost identical rotational acquisitions is performed at different heart phases to obtain a complete data set at a minimum temporal resolution of 1/N of the heart cycle time. First results in simulation, using an experimental phantom, and in preclinical in vivo studies showed that excellent image quality can be achieved

Towards cardiac angiographic computed tomography

Minimal invasive, intravascular, cardiac therapy would tremendously benefit from sophisticated image guidance. Currently tomographic imaging is available only offline by pre therapeutic CT scans. Recent developments in angiographic computed tomography (ACT) allow 3D imaging of low contrast details with angiographic, C-arm based X-ray devices typically used during intervention. To take advantage of the improved contrast resolution a new approach for cardiac imaging is proposed which is based on intravenous contrast injection. The method is an analogue to the multi-segment reconstruction strategies in cardiac spiral CT. The challenge is to exclude movement in the reconstruction process as much as possible. Under the assumption of an almost periodic movement of the heart only those projection data are used for reconstruction which are acquired at a heart phase where the heart is expected to be almost in rest. Data gaps due to ECG-gating are filled by a series of N temporally complementary but spatially redundant scans. Temporal resolution is increased to 1/N of the heart cycle time. Intravenous contrast injection allows for longer acquisition time to perform the series of scans. First results in simulation studies and in experimental phantom studies showed that excellent image quality could be achieved with the new method.