|
| |
|
|
Detection of foreign objects in X-ray chest images using machine learning methods
Matoušková, Barbora ; Kolář, Radim (referee) ; Chmelík, Jiří (advisor)
Foreign objects in Chest X-ray (CXR) cause complications during automatic image processing. To prevent errors caused by these foreign objects, it is necessary to automatically find them and ommit them in the analysis. These are mainly buttons, jewellery, implants, wires and tubes. At the same time, finding pacemakers and other placed devices can help with automatic processing. The aim of this work was to design a method for the detection of foreign objects in CXR. For this task, Faster R-CNN method with a pre-trained ResNet50 network for feature extraction was chosen which was trained on 4 000 images and lately tested on 1 000 images from a publicly available database. After finding the optimal learning parameters, it was managed to train the network, which achieves 75% precision, 77% recall and 76% F1 score. However, a certain part of the error is formed by non-uniform annotations of objects in the data because not all annotated foreign objects are located in the lung area, as stated in the description.
|
|
|
Processing of X-Ray images in studying jawbone diseases
Kabrda, Miroslav ; Šmirg, Ondřej (referee) ; Mikulka, Jan (advisor)
The subject of this thesis is a method proposed for automated evaluation of the parameters of X-ray of cystic disorders in human jawbones. The main problem in medical diagnostic is the low repeatability due to the subjective evaluation of images without using a tool for image processing. In this thesis are described the basic steps of image processing, various methods of image segmentation and chosen segmentation method live-wire. Selected segments were processed in the ImageJ Java environment. In the cystic regions their basic statistical and shape properties were evaluated. The obtained values were used for learning the classification model (decision tree) in the environment RapidMiner. This model was used to create a plug-in for automatic classification of the type of cysts in the program ImageJ.
|
|
|
XRD characterization of electrodes and electrode materials for electrochemical power sources
Klvač, Ondřej ; Chladil, Ladislav (referee) ; Čech, Ondřej (advisor)
The thesis describes available choices of methods for research of electrochemical power source electrode strcture by x-ray diffraction spectroscopy. At first there is a theoretical computation of difractogram. Then there are discussed possibilities of in-situ and ex-situ techniques as background research. Practical part of thesis compares theoretical assumptions to measurements of lithium and sodium cells, including analyze of crystallographic phases and lattice constants. As a conclusion, there are discussed possibilities of subsequent research.
|
|
|
Precise setup of production process to minimalize presence of voids in solder joints where were used alloy SAC305
Slavík, Pavel ; Macho, Tomáš (referee) ; Valach, Soběslav (advisor)
The theoretical part of dissertation is devoted to familiarization with the process of assembly and soldering of PCBs. There are described the selected types of lead-free solder paste, the various procedures of flux deposition, where the greater part is focuses to the printing through template. Also, I analyze the different procedures shouldering and soldering. I focused mainly on the mechanical mounting of SMD components soldering with help convection reflow and soldering in the vapor. In the practical part is described the design of the test PCB, methodology of assembly PCB and soldering various technologies with predefined settings. The analysis of the results is done by a non-destructive x-ray method and destructive micro-cut method. The x-ray method detects the position and size of the voids in the solder. This information is used for faster processing of micro-cut. The micro-cut method is included in the analysis for a more detailed examination of the solder joint. At the end of the thesis, there is a summary of the recommendation that the setting of the manufacturing process increases the quality of the brazed joint.
|
|
|
Deep Learning for Medical Image Analysis
Dronzeková, Michaela ; Kodym, Oldřich (referee) ; Španěl, Michal (advisor)
The purpose of this thesis is to use convolutional neural networks for X-ray image classification of human body. Four different architectures of neural networks have been created. They were trained and tested on three tasks: classification of front and lateral chest, classification of X-ray images into several different categories and classification of diseases in chest X-ray. ResNet and SEResNet architectures achieved the best results. SEResNet scored 99,49% accuracy in the first task, ResNet achieved 94,97% accuracy in the second task and SEResNet reached 31,53% in the third task with F1 measure as metrics for evaluating results.
|
|
|
Detection, Extraction and Measurement of the Length of the Metacarpal Bones in X-Rays of the Human Hand
Chalupnik, Alexandr ; Rydlo, Štěpán (referee) ; Drahanský, Martin (advisor)
The aim of the bachelor thesis is the design and implementation of an algorithm for the detection, extraction and measurement of the length of the metacarpal bones in x-rays of the human hand. The first part is focused on the principles of x-rays and the anatomy of the human hand. The second part contains the design and implementation of the task. The algorithm is based on image processing methods, such as thresholding or edge detection. In the end, the measurement results are evaluated on a set of images, which were provided by the Department of Anthropology at Masaryk University.
|
|
|
Design of metal detector and X-ray luggage scanner
Fridrichová, Eva ; Paliatka, Peter (referee) ; Zvonek, Miroslav (advisor)
Subject of this Diploma thesis is the design of metal detector and luggage X-ray screening. The thesis concerns with design of frame metal detector and luggage x-ray with focus on visual appeal, keeping of functional elements and concept of construction solution. Main objective is to find solution how to connect aesthetic, ergonomic and technical aspect. Complete analysis leads to design of metal detector and luggage X-ray screening. These appliances then appeal as compact system with uniform design. Also the machines can be used separately. Design of this system of appliances is solved through ergonomic, psychologic, technical and aesthetic aspects.
|
|
|
Spectral computed tomography: comparison of real and virtual nativ images
Širůčková, Kateřina ; Bartušek, Karel (referee) ; Marcoň, Petr (advisor)
Advanced X-ray medical diagnostic methods massively increase the radiation exposure of patients. Therefore, it is necessary to focus on the reduction of radiation and its stochastic and deterministic effects. This project compares real native images TNC with virtual native images VNC. Virtual native images are acquired by spectral computed tomography method and it is suggested that VNC could potentially substitute real native images, thereby, the total radiation dose from multiphase spectral CT would decrease. A comparison was performed by defining certain parameters that represented the differentiation of the measured and calculated values in the images. The parameters were SNR, CNR, absolute difference, relative error and they were analysed by statistical tests using p-value and correlation analysis. Another output is a database of patients with compared parameters from real and virtual native images.
|
|
|
Analysis of Defects on PCB Using Modern Optical Method
Vala, Martin ; Řezníček, Michal (referee) ; Řihák, Pavel (advisor)
This diploma thesis is focused on detecting of defects on BGA (Ball Grid Array) components using of X-ray. Defects are formed during reflow BGA components during assembly, but also later due to mechanical and thermal stress. Therefore, there is an overview of defects and methods of diagnosis of BGA packages eg .: modern X-ray defect detection or micro sections. There is disclosed a device NORDSON DAGE XD7600NT its operation and setup. The device enables advanced methods of scanning called X-plane. For creating 3D models used reconstructive software called CERA, which uses raw data from the method of X-Plane.
|
guest ::
