|
|
Deep-learning-based pattern detection in medical images
Koščová, Zuzana ; Vičar, Tomáš (referee) ; Jakubíček, Roman (advisor)
This Bachelor thesis deals with Deep-learning-based pattern detection in medical images. For better understanding of a subject artificial neural network and convolutional neural network (CNN) are described at first. Next chapter is focused on specific detection methods which use CNN. Within a bachelor thesis a dataset of abdominal CT a MRI scans was created. Faster R-CNN and YOLO algorithms were trained and tested on acquired scans for liver detection. Implementation of chosen methods took place in Python programming language using the Pytorch library. Finally, detection results and possible use in medicine are discussed.
|
|
|
Segmentation of cartilage tissue of mouse embryos in 3D micro CT data
Matula, Jan ; Vičar, Tomáš (referee) ; Chmelík, Jiří (advisor)
Manual segmentation of cartilage tissue in micro CT images of mouse embryos is a very time consuming process and significantly increases the time required for the research of mammal facial structure development. This problem might be solved by using a fully-automatic segmentation algorithm. In this diploma thesis a fully-automatic segmentation method is proposed using a convolutional neural network trained on manually segmented data. The architecture of the proposed convolutional network is based on the U-Net architecture with it's encoding part substituted for the encoding part of the VGG16 classification convolutional neural network pretrained on the ImageNet database of labeled images. The proposed network achieves Dice coefficient 0.8731 ± 0.0326 in comparison to manually segmented images.
|
|
|
Registration of image sequences from experimental video-ophthalmoscope
Bjelová, Martina ; Vičar, Tomáš (referee) ; Kolář, Radim (advisor)
The topic of this thesis is registration of image sequences captured by experimental ophthalmoscope. It contains anatomical description of the visual system as well as the description of functions of selected ophthalmoscopic devices. The next covered topic is theoretical summary of registration process, which is followed by an overview of the used methods, which forms the basis of the design and implementation of the registration algorithm in the Python programming language. After implementation, the accuracy and computational complexity of a registration was evaluated. Tests of optimalization of the proposed approach were performed with regards to the obtained results, through which sufficiently accurate registration has been achieved, evaluated on the basis of Euclidean distances, standard deviation and visual observation. In case of high-quality recorded sequences, values of Euclidean distances ranged from 0.60 to 4.07 pixels on the contrary, values higher than 20 pixels occurred in the case of poor-quality recordings. Standard deviation values in recordings with high enough resolution have not reached worse results than 4.12.
|
|
|
Suppression of the responsive component of electrodermal activity
Vraný, Jakub ; Vičar, Tomáš (referee) ; Kolářová, Jana (advisor)
Electrodermal acitivity is a kind of electrochemical signal generated with relation to activity of the autonomic nervous system that stimulates the sweat glands. In this way, is it possible to measure the activity of the sympathetic part of the nerve systém and evaluate the cognitive stress of the treated person, which is manifested by responsive signals in EDA record, respectively to increased occurence of responses. The aim of this work is to design a deep learning algorithm for the identification of this component in the record of data taken from UBMI database. The recordings contain a sequence of measurements the conductance of the skin of patient, who was subjected alternately to the states of rest and subsequently a state of mental stress. The data were annotated according to presence of the responsive components occuring in the records of EDA. Subsequently, a suitable deep learning algorithm was implemented in order to classify the responsive components in the measured EDA signal. The neural network model has been taught, optimized and implemented on the measurement samples using annotated data. The obtained results data were statistically evaluated to qualify the success of the classification of responsive components and differences in the records of mental calm and stress. The results of the classification and comparison of EDA records measured at different conditions of the patient were discussed subsequently.
|
|
|
Artificial intelligence for predicting sepsis from clinical signals
Šidlo, David ; Chmelík, Jiří (referee) ; Vičar, Tomáš (advisor)
This bachelor thesis deals with the issue of predicting sepsis from clinical data using artificial intelligence methods. In the theoretical part, a literature research is made on the basic principles and functioning of various methods of artificial intelligence. Greater emphasis was placed on recurrent neural networks. The aim of the practical part was to implement a suitable method in the chosen programming environment. The LSTM network and the temporal convolutional network TCN were chosen as suitable methods. The best results of the normalized value of the utility score were achieved by TCN, namely 0.377 and seven-layer LSTM 0.356.
|
|
|
Segmentation of kidney tumor in CT data
Urbanová, Hedvika ; Vičar, Tomáš (referee) ; Jakubíček, Roman (advisor)
This diploma thesis deals with the kidney tumor segmentation in CT data. First kidney anatomy and pathology is discussed. Following topics are the conventional segmentation techniques and segmentation techniques using machine learning. In the final part, the convolutional neural network is discussed as its algoritm was used for segmentation in the practical part, in which algoritm for segmentation was designed in Python programming language. This algoritm was tested and evaluated using databaze KiTS19.
|
|
|
Monitoring system for detecting the motility and position of laboratory animals after anesthesia
Enikeev, Amir ; Vičar, Tomáš (referee) ; Čmiel, Vratislav (advisor)
This diplom work entitled "Monitoring system for the detection of motility and position of laboratory animals after anesthesia" focuses on the design and implementation of non-contact detection of the rat or mouse position in the enclosure with a transparent cover. The aim of this semester paper is to find suitable methods of realizing contactless detection of the position of a laboratory rat or mouse. This automatic positioning of the animal will be used as the basis for controlling the irradiator in the next follow-up work, which will "shade" animal movement and aim at the scar on the animal's back. The rat that is located inside our enclosure is either standard or movable after anesthesia. In this work I first deal with searches of automatic monitoring systems for detecting the position of animals in the enclosure. Then, in the practical part, I test three types of cameras for image detection of rat position. Evaluation software for motion analysis will largely be solved in the follow-up diploma thesis.Project made like monitoring and detecting software based on OpenCV.
|
|
|
Segmentation of biological samples in cryo-electron microscopy images using machine learning methods
Sokol, Norbert ; Vičar, Tomáš (referee) ; Chmelík, Jiří (advisor)
Zobrazovanie pomocou kryo-elektrónovej mikroskopie má svoje nezastúpiteľné miesto v analýze viacerých biologických štruktúr. Lokalizácia buniek kultivovaných na mriežke a ich segmentácia voči pozadiu alebo kontaminácii je základom. Spolu s vývojom viacerých metód hlbokého učenia sa podstatne zvýšila úspešnosť úloh sémantickej segmentácie. V tejto práci vyvinieme hlbokú konvolučnú neurónovú sieť pre úlohu sémantickej segmentácie buniek kultivovaných na mriežke. Dátový súbor pre túto prácu bol vytvorený pomocou dual-beam kryo-elektónového mikroskopu vyvinutého spoločnosťou Thermo Fisher Scientific Brno.
|
|
|
Module for Electrodermal Activity recording
Vičar, Tomáš ; Harabiš, Vratislav (referee) ; Bubník, Karel (advisor)
This thesis describes electrodermal activity (EDA) and its origin based on the properties of the skin and thermoregulation of body. EDA is a signal having a close relationship to psychophysiology and its help we can evaluate a variety of emotional, motoric and attentional effects on the human organism. The thesis also discusses the possibility of sensing skin potential and conductatce and how to construct a module for its scanning and uploading to computer.
|
|
|
Utilization of convolutional neural networks for segmentation of mouse embryos cartilaginous tissue in micro-CT data
Poláková, Veronika ; Vičar, Tomáš (referee) ; Chmelík, Jiří (advisor)
Automatická segmentace biologických struktur v mikro-CT datech je stále výzvou, protože často objekt zájmu (v našem případě obličejová chrupavka) není charakterizovaný unikátním jasem či ostrými hranicemi. V posledních letech se konvoluční neuronové sítě (CNNs) staly mimořádně populárními v mnoha oblastech počítačového vidění. Konkrétně pro segmentaci biomedicínských obrazů je široce používaná architektura U-Net. Nicméně v případě mikro-CT dat vyvstává otázka, zda by nebylo výhodnější použít 3D CNN. Diplomová práce navrhla CNN architekturu založenou na síti V-Net včetně metodologie pro předzpracování a postprocessing dat. Základní architektura byla dále optimalizována pomocí pokročilých architektonických modifikací jako jsou pyramidální modul dilatovaných konvolucí (ASPP modul), škálovatelná exponenciálně-lineární jednotka (SELU aktivační funkce), víceúrovňová kontrola učení (multi-output supervision) či bloky s hustými propojeními (Dense blocks). Pro učení sítě byly použity moderní přístupy jako zahřívání kroku učení (learning rate warmup) či AdamW optimalizátor. I přes to, že 3D CNN v úloze segmentace obličejové chrupavky nepřekonala U-Net, optimalizace zvýšila medián Dice koeficientu z 69,74 % na 80,01 %. Používání těchto pokročilých architektonických modifikací v dalším výzkumu je proto vřele doporučováno, jelikož můžou být přidány do libovolné architektury typu U-Net a zároveň výrazně zlepšit výsledky.
|
guest ::
