|
|
Automatic AIF voxel detection in DCE-MRI using machine learning
Frolíková, Štěpánka ; Jiřík, Radovan (referee) ; Vitouš, Jiří (advisor)
Dynamická kontrastní magnetická rezonance (DCE-MRI) je metoda používaná k hodnocení perfuze tkání. Vyžaduje co nejpřesnější identifikaci a detekci arteriální vstupní funkce (AIF). Manuální značení AIF není dostatečně rychlé a může být dost pracné. Obvykle se ale algoritmy automatické detekce nepoužívají. Cílem této práce je přesné určení AIF pomocí natrénovaného modelu využívající strojového učení. Výsledky natrénovaného modelu jsou porovnány s jinými přístupy, jako je k-means shlukování. Data použitá pro trénování a testování modelu jsou jak reálná, tak syntetická. Syntetická data jsou simulována pomocí farmakokinetického modelu DCATH. Shlukovací model je založen na algoritmu k-means, který je optimalizován pro lidská a myší data. Model strojového učení využívá klasifikátor založený na metodě náhodného lesa v kombinaci se shlukováním. Tato práce hodnotí přesnost použitých metod, jejich výhody a nevýhody. Plně automatická detekce AIF by značně ušetřila délku trvání perfúzní analýzy a kvalitu odhadu AIF.
|
|
|
Simulation of contrast-agent distribution in a real microvascular system
Schwachulová, Natalia ; Mézl, Martin (referee) ; Jiřík, Radovan (advisor)
The goal of this thesis is to simulate the distribution of contrast agents in a real microvascular system in order to validate methods used in perfusion imagining in DCE-MRI. It contains explanation of principles of the DCE-MRI method and introduction of blood vessels in each of the murine tissue datasets. The thesis resolves around finding graph model, representing blood vessels in used datasets, calculating diameter and respective vascular resistance of each blood vessel, determining direction of blood flow in the blood territory, estimation of impulse residual function of each tissue and creation of generator of concentration curves. Obtained concentration curves are then used to estimate perfusion parameters , and and analyzed using SW PerfLab in order to determine, which pharmacokinetic model is the best to represent each tissue.
|
|
|
DEEP LEARNING FOR SINGLE-VOXEL AND MULTIDIMENSIONAL MR-SPECTROSCOPIC SIGNAL QUANTIFICATION, AND ITS COMPARISON WITH NONLINEAR LEAST-SQUARES FITTING
Shamaei, Amirmohammad ; Latta,, Peter (referee) ; Kozubek, Michal (referee) ; Jiřík, Radovan (advisor)
Pro získání koncentrace metabolitů ve vyšetřované tkáni ze signálů magnetické rezonanční spektroskopie (MRS) je nezbytné provézt předzpracování, analýzu a kvantifikaci MRS signálu. Rychlý, přesný a účinný proces zpracování (předzpracování, analýza a kvantifikace) MRS dat je však náročný. Tato práce představuje nové přístupy pro předzpracování, analýzu a kvantifikaci MRS dat založené na hlubokém učení (DL). Navržené metody potvrdily schopnost použití DL pro robustní předzpracování dat, rychlou a efektivní kvantifikaci MR spekter, odhad koncentrací metabolitů in vivo a odhad nejistoty kvantifikace. Navržené přístupy výrazně zlepšily rychlost předzpracování a kvantifikace MRS signálu a prokázaly možnost použití DL bez učitele. Z hlediska přesnosti byly získány výsledky srovnatelné s tradičními metodami. Dále byl zaveden standardní formát dat, který usnadňuje sdílení dat mezi výzkumnými skupinami pro aplikace umělé inteligence. Výsledky této studie naznačují, že navrhované přístupy založené na DL mají potenciál zlepšit přesnost a efektivitu zpracování MRS dat pro lékařskou diagnostiku. Disertační práce je rozdělena do čtyř částí: úvodu, přehledu současného stavu výzkumu, shrnutí cílů a úkolů a souboru publikací, které představují autorův přínos v oblasti aplikací DL v MRS.
|
|
|
Projection and projection-reconstruction x-ray imaging process simulation
Fiala, Petr ; Jiřík, Radovan (referee) ; Drastich, Aleš (advisor)
The work deals with physical principles of X-ray generation and development of image during projection and projection reconstruction. A proposal of user’s application in a Matlab – Guide is given, which can be used as a laboratory exercise of the simulation of the projection- and projection image reconstruction. The computer program involves an evaluation of a X-ray quality of CT RTG ZS – quantitative assessment of spatial resolution and as well as the acquisition contrast as a function on an object size. The main aim of the work was the comparison of the acquisition contrast at various acquisition projection and projection-reconstruction parameters. Also, the work is illustrated by some results achieved.
|
|
|
Modeling of Respiratory System
Matoušek, Josef ; Roleček, Jiří (referee) ; Jiřík, Radovan (advisor)
This thesis is concerned with respiration system simulation. The main goal of this studies is to gain knowledge of how human respiration system works, along with selecting proper algorithms and their implementation into mathematic model. These models are realized by Simulink interface, which is part of Matlab environment. Similar models are used for medical technology research, especially for artificial ventilation of the lungs. There is a practical utilisation of these models suitable for illustration and respiration system's biological processes simulation, particularly combined model described in GUIDE graphic interface - also implemented in Matlab environment.
|
|
|
Experimental ultrasonic measurement using LabView
Baše, Michal ; Jiřík, Radovan (referee) ; Kolář, Radim (advisor)
This bachelor´s thesis deals with ultrasonic field, measurement of its attenuation and measurement using LabVIEW. There are explained basic parameters of ultrasonic transducers and the ultrasonic field. Two methods for measuring ultrasonic attenuation and modification of the second one were proposed. There are introduced two programs in the LabVIEW for attenuation measurement, first one generates continual signal, which is brought to transducer and second one generates pulse of a few periods of sine. At the end there are showed produce of measurement using these programs.
|
|
|
Flow Measurement Using Dilution Methods
Pokorný, Pavel ; Jiřík, Radovan (referee) ; Harabiš, Vratislav (advisor)
The bachelor thesis focuses on the construction of a program in LabVIEW to measure pH with the Vernier sensing probe connected through the interface LabPro and to be able to inset the measured data with a physical model of a dilution curve to get information about the model. It further describes the necessary requirements of the product in which the measurement of the dilution curve would be possible, the principle of constructing it and its usage with the above mentioned program to construct a laboratory task.
|
|
|
Optimization of Contrast in MRI of Labeled Thrombi
Dohnalová, Petra ; Mézl, Martin (referee) ; Jiřík, Radovan (advisor)
This bachelor thesis deals with the general principle of molecular MRI, the formation, composition of trombi and their impact on the human body. It also deals with the contrast agents, to which molecules are they targeted and why, as well as the principle of interactions with these target molecules. The following are found values of native relaxation times for thrombi, blood, gray and white matter and relaxivities of contrast agents at different values of B0 magnetic field. The next part is dedicated to the program for simulations of T1- and T2- weighted MRI and searching for the optimal setting of acquisition parametersbased on the achieved contrast. Finally, the possibility of combining T1- and T2- weighted images to increase contrast is explored.
|
|
|
Modelling in perfusion MR imaging
Válková, Hana ; Jiřík, Radovan (referee) ; Kratochvíla, Jiří (advisor)
This thesis deals with the magnetic resonance perfusion data analysis especially DCEMRI. In its introduction the thesis describes the problem of DCE-MRI data aquisition, the necessity of appropriate contrast agent and basic principles of perfusion analysis. The dynamic behavior of contrast agent vascular distribution can be described by arterial input function (AIF). The shape of the curves close to the area of interest is affected by dispersion which is called vascular transport function (VTF) due to the distribution of the contrast agent to the region of interest. Finally the tissue residual function describes system behavior of tissue. The practical part of the diploma thesis is aimed at implementation of model curves AIF, VTF and TRF. Furthermore, a simulation program was created for easy manipulation with introduced models moreover the program is used to perform an estimation of perfusion parameters based on nonblind deconvolution. The method is validated on synthetic data and illustrated on clinical data of the renal cell carcinoma patient.
|
|
|
Modelling of parametric arterial input function in perfusion imaging
Černý, Štěpán ; Jiřík, Radovan (referee) ; Kratochvíla, Jiří (advisor)
The task of this bachelor’s tesis is to study the subject in the area of perfusion imaging based on dynamic imaging with T1 contrast. Aims is primarily on the role of the arterial input function. Principles phenomenon of dynamic magnetic resonance measurements are discribed. The role of contrast media and their influence on the relaxation time T1 are described. It is also described problems perfusion analysis, measurement and mathematical modeling of arterial input function and perfusion acquisition parameters. In the experimental part, there is modeled several models of arterial input function and their evaluation. Further the creation program for the approximation real measurement of arterial input function created models and the integration of this program into a graphical interface created for work with models. Was also verified the functionality of the program and evaluate the success of approximations for individual models.
|
guest ::
RSS feed.