Ray-based approach to skeletal muscle perfusion measurement on interventional x-ray systems

Abstract

Periprocedural assessment of tissue perfusion by imaging methods could improve outcome control during treatment of peripheral vascular disease. Currently, endovascular revascularization treatments are assessed by planar angiography which only allows for qualitative inspection of blood flow in vessels. In this paper, we present a method for periprocedural perfusion estimation based on temporal attenuation curves in skeletal muscles using angiographic C-arm systems. The proposed method tackles the loss of spatial depth information which occurs in conventional angiography by combining the acquired angiograms with two additional C-arm rotational soft tissue scans. The area subject to contrast propagation is segmented from the two images that are tomographically reconstructed from the rotational scans and is then used to estimate the spatially averaged temporal contrast attenuation along the x-ray directions from the angiograms. A segmentation method which is tailored to the estimation procedure is applied to limit inaccuracies in the estimation. The accuracy of the method in estimating tissue blood flow in muscular tissue is evaluated in a simulation study using experimental data from CT perfusion acquisitions. Results show that perfusion estimation accuracy is limited owing to spatial inhomogeneity of contrast in muscular tissue and to the presence of vessels along the x-ray directions. Nonetheless, the spatially averaged perfusion quantification allows for improved visual differentiation of normal and hypoperfused tissue in comparison with conventional digital subtraction angiography. Periprocedural assessment of muscle perfusion through digital subtraction angiography is challenging due to lack of longitudinal information in the planar projections. By including additional 3D information on the anatomy retrieved from rotational soft tissue scans, the visualization and differentiation of normal and hypoperfused areas can be improved.