University of Twente Student Theses


2D perfusion angiography in patients with critial limb ischemia: a study into reproducibility, variability and vessel extraction

Janssens, A.J.B. (2018) 2D perfusion angiography in patients with critial limb ischemia: a study into reproducibility, variability and vessel extraction.

[img] PDF
Abstract:Critical limb ischemia (CLI) is a serious condition caused by end stage peripheral arterial disease in which the viability of the limb is at risk. Endovascular therapy has become the treatment of choice, because of its minimal invasive nature and high technical success rate. The aim of revascularization is to increase oxygen supply to the wound so that tissue perfusion, oxygen and nutrient levels become sufficiently high to enable wound healing. Unfortunately, the success of endovascular revascularizations is largely unpredictable, as demonstrated by the high rates of delayed wound healing and repeated interventions. Therefore, an objective method to assess the adequacy of revascularization and subsequent tissue perfusion following endovascular treatment is needed. 2D perfusion angiography (2DPA) is a post processing technique which uses digital subtraction angiography (DSA) images to quantify density changes caused by propagation of contrast medium. Within a certain observer determined region of interest (ROI), pre- and post-interventional density change over time can be visualized as time density curves (TDC). The currently available literature supports the feasibility of 2DPA in CLI patients, because it enables quantification and comparison of pre- and post-interventional perfusion results. Several important aspects need to be researched further to enable and optimize clinical use of 2DPA, among others research into reproducibility, separation of the macro- and microcirculation to enable assessment of tissue perfusion alone, and reduction of the influence of motion artifacts. This thesis aims to study the first two, namely reproducibility and vessel separation. Because literature is lacking on reproducibility and observer agreement, both are studied in this thesis. In eleven patients, reproducibility was investigated by comparison of two perfusion acquisitions, which were obtained under equal patient and acquisition settings. Initial reproducibility was only 64%. Reproducibility of three out four unreproducible TDC pairs was achieved by removal of the frames before contrast arrival, and inclusion of the same number of frames. This resulted in a reproducibility of 91%. Optimal analyzing conditions and the effect on reproducibility need to be researched further. To determine observer variability, two ROIs were researched, one including the complete foot with exclusion of the digits, and one limited to the wound area. For intra-observer variability, one observer drew each ROI five times in each of the ten included acquisitions. For inter-observer variability, two observers drew each ROI in twenty acquisitions. Both analyses showed excellent agreement for the ROI including the foot. Variability of the ROI including the wound area was high, because interpretation of the wound area and location appeared to differ within and between observers. Therefore, using the ROI including the foot is preferred over the ROI limited to the wound area. To improve tissue perfusion assessment using the ROI of the whole foot, a method to extract the arteries from perfusion images was presented. Successful separation resulted in a typical microvascular and arterial perfusion curve, which was demonstrated by two clinical cases. Clinical validation studies must confirm the added value of this extraction method.
Item Type:Essay (Master)
St. Antonius Hospital, Nieuwegein, The Netherlands
Faculty:TNW: Science and Technology
Subject:44 medicine
Programme:Technical Medicine MSc (60033)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page