|
|
Statistical properties of ultrasound images in contrast mode
Věžníková, Romana ; Kolář, Radim (referee) ; Slávik, Vladimír (advisor)
The bachelor’s thesis is focused on properties of ultrasound images in contrast mode. First of all it describes the basic properties of ultrasound, the way in which the ultrasonic signals are generated, the parameters of ultrasound and ultrasonic imaging modes. Next part deals with harmonic and contrast imaging modes. This part explains principles of these modes as well as the usage of contrast agents and artefacts connected with ultrasound imaging in contrast mode. The thesis also shows the approximation of the amplitude probability density function of ultrasonic images and description of its single models, out of which is chosen the RiIG model (Rician Inverse Gaussian distribution) for approximation of ultrasound images in this case. To gain parameters of amplitude probability density function from ultrasonic data in contrast imaging mode is made the application using Matlab programming environment. Thereafter is the thesis focused on determining equations for dependences of RiIG distribution parameters on known concentrations of contrast agent for ultrasonic data captured in vitro. These equations are useful for determining the unknown concentration of contrast agent used in images captured in vivo and making its dilution function.
|
|
|
Updating the capacity model of roundabouts
Patočka, Miroslav ; Bačová, Katarina (referee) ; Kumpošt, Petr (referee) ; Radimský, Michal (advisor)
The doctoral thesis deals with a new approach to the characteristics of traffic flows at roundabouts in relation to the capacity estimation. The currently used methodology, which was introduced in 1973 by Werner Siegloch, is based on imprecise assumptions and simplifications, which may ultimately distort the results of capacity assessments carried out according to the existing valid national regulation. This thesis uses not only an innovative procedure for obtaining empirical data by image analysis using so-called computer vision, but above all, the process of gap-acceptance of main-stream time clearances by minor stream vehicles is solved as a stochastic phenomenon. This approach aims to replace the deterministic values of the critical gap and follow-up headway with an appropriate probability density function that will meet empirical data in the best way.
|
|
|
Statistical properties of ultrasound images in contrast mode
Věžníková, Romana ; Kolář, Radim (referee) ; Slávik, Vladimír (advisor)
The bachelor’s thesis is focused on properties of ultrasound images in contrast mode. First of all it describes the basic properties of ultrasound, the way in which the ultrasonic signals are generated, the parameters of ultrasound and ultrasonic imaging modes. Next part deals with harmonic and contrast imaging modes. This part explains principles of these modes as well as the usage of contrast agents and artefacts connected with ultrasound imaging in contrast mode. The thesis also shows the approximation of the amplitude probability density function of ultrasonic images and description of its single models, out of which is chosen the RiIG model (Rician Inverse Gaussian distribution) for approximation of ultrasound images in this case. To gain parameters of amplitude probability density function from ultrasonic data in contrast imaging mode is made the application using Matlab programming environment. Thereafter is the thesis focused on determining equations for dependences of RiIG distribution parameters on known concentrations of contrast agent for ultrasonic data captured in vitro. These equations are useful for determining the unknown concentration of contrast agent used in images captured in vivo and making its dilution function.
|
|
| |
|
| |
|
|
Pay-per-bid Auctions
Bocák, Petr ; Svoboda, Miroslav (advisor) ; Bolcha, Peter (referee)
This paper analyzes pay-per-bid auctions which have appeared recently on the Internet. In this paper I test the suitability of Platt, Price and Tappen's model using data about ended auctions on the website Bonus.cz. The model forecasts the distribution of ending prices depending on the item's value, bid fee and price increment. Observed ending prices distribution of more than 70% of commonly auctioned items on Bonus.cz fit the model. Furthermore, I discuss the relation between the time of day and the number of ended auctions and the influence the price increment has on the collected revenue. Contrary to model predictions, I observed that auctions with the price increment of 0,1 CZK generated significantly higher revenue than auctions with price increment of 1CZK.
|
guest ::
