Development of an in vitro model for studying aneurysms

The aims of this project was to develop an arterial aneurysm using either enzymatic or laser degradation of the arterial wall without affecting the viability of the tissue and to cultivate the arteries under pulsatile flow conditions in a vascular bioreactor with a view to investigate the progress of the disease. Characteristics of aneurysms are the degradation of smooth muscle cells, collagen and elastin. Detached smooth muscle cells and degradation of the collagen matrix and elastin fibres were observed in arteries degraded with enzymes elastase and collagenase. Only remnants of the arterial wall were detected after cultivation. This might be a suitable model for late stage aneurysms. Arteries treated with the laser system showed no charring or heat damage of the not dissected area. Collagen matrix, smooth muscle cells and elastin fibres were intact. A clear defined cut was made in a depth of 200 μm and tissue was removed. Following cultivation of these arteries a dilation of the laser-eroded area was observed. This model can mimic atherosclerotic aneurysms, when plaques weaken the tunica media of the blood vessel wall and rupture. Limitations of this study were contamination of the bioreactor system and a low number of cultivations. The aim to generate a living arterial aneurysm in vitro was not achieved. Tissue viability decreased to the level of negative controls after cultivation.

Application of the Hidden Markov Model for Innovative Projects "Viability" Analysis

This master thesis deals with determining of innovative projects "viability". "Viability" is the probability of innovative project being implemented. Hidden Markov Models are used for evaluation of this factor. The problem of determining parameters of model, which produce given data sequence with the highest probability, are solving in this research. Data about innovative projects contained in reports of Russian programs "UMNIK", "START" and additional data obtained during study are used as input data for determining of model parameters. The Baum-Welch algorithm which is one implementation of expectation-maximization algorithm is used at this research for calculating model parameters. At the end part of the master thesis mathematical basics for practical implementation are given (in particular mathematical description of the algorithm and implementation methods for Markov models).