Perfusion magnetic resonance imaging is modern diagnostic method used mainly in oncology. In this method, contrast agent is injected to the subject and then is continuously monitored the progress of its concentration in the affected area in time. Correct determination of the arterial input function (AIF) is very important for perfusion analysis. One possibility is to model AIF by multichannel blind deconvolution but the estimated AIF is necessary to be scaled. This master´s thesis is focused on description of scaling methods and their influence on perfussion parameters in dependence on used model of AIF in different tissues.

This thesis deals with the possibilities of determining perfusion parameters of vascular system, using contrast-enhanced ultrasound imaging, which is non-invasive method. Properties of ultrasonography and use of contrast agents are briefly summarized. The methods selected for perfusions analysis were Bolus-tracking¬¬, Burst-replenishment and both of them combined – Bolus&Burst. Parametric models based on these methods were created for modelling an approximation of set perfusion parameters with the use of blind deconvolution.

This master thesis deals with the contrast agents and their application in the ultrasound perfusion analysis. It is focused on Bolus & Burst method which, as a combination of two approaches that have been used so far, allows an absolute quantification of perfusion parameters in the region of interest. Contrast agent concentration time sequence is modeled as a convolution of the parametrically defined arterial input function and the tissue residual funkction. Thesis discusses different mathematical models of these functions as well as the methods of the parameters estimation. The methods functionality is validated on simulated and also preclinical data.

This Master Thesis deals with image restoration using deconvolution. The terms introducing into deconvolution theory like two-dimensional signal, distortion model, noise and convolution are explained in the first part of thesis. The second part deals with deconvolution methods via utilization of the Bayes approach which is based on the probability principle. The third part is focused on the Alternating Minimization Algorithm for Multichannel Blind Deconvolution. At the end this algorithm is written in Matlab with utilization of the NAG C Library. Then comparison of different optimization methods follows (simplex, steepest descent, quasi-Newton), regularization forms (Tichonov, Total Variation) and other parameters used by this deconvolution algorithm.