|
| |
|
|
Segmentation of multiple sclerosis lesions using deep neural networks
Sasko, Dominik ; Myška, Vojtěch (referee) ; Kolařík, Martin (advisor)
Hlavným zámerom tejto diplomovej práce bola automatická segmentácia lézií sklerózy multiplex na snímkoch MRI. V rámci práce boli otestované najnovšie metódy segmentácie s využitím hlbokých neurónových sietí a porovnané prístupy inicializácie váh sietí pomocou preneseného učenia (transfer learning) a samoriadeného učenia (self-supervised learning). Samotný problém automatickej segmentácie lézií sklerózy multiplex je veľmi náročný, a to primárne kvôli vysokej nevyváženosti datasetu (skeny mozgov zvyčajne obsahujú len malé množstvo poškodeného tkaniva). Ďalšou výzvou je manuálna anotácia týchto lézií, nakoľko dvaja rozdielni doktori môžu označiť iné časti mozgu ako poškodené a hodnota Dice Coefficient týchto anotácií je približne 0,86. Možnosť zjednodušenia procesu anotovania lézií automatizáciou by mohlo zlepšiť výpočet množstva lézií, čo by mohlo viesť k zlepšeniu diagnostiky individuálnych pacientov. Našim cieľom bolo navrhnutie dvoch techník využívajúcich transfer learning na predtrénovanie váh, ktoré by neskôr mohli zlepšiť výsledky terajších segmentačných modelov. Teoretická časť opisuje rozdelenie umelej inteligencie, strojového učenia a hlbokých neurónových sietí a ich využitie pri segmentácii obrazu. Následne je popísaná skleróza multiplex, jej typy, symptómy, diagnostika a liečba. Praktická časť začína predspracovaním dát. Najprv boli skeny mozgu upravené na rovnaké rozlíšenie s rovnakou veľkosťou voxelu. Dôvodom tejto úpravy bolo využitie troch odlišných datasetov, v ktorých boli skeny vytvárané rozličnými prístrojmi od rôznych výrobcov. Jeden dataset taktiež obsahoval lebku, a tak bolo nutné jej odstránenie pomocou nástroju FSL pre ponechanie samotného mozgu pacienta. Využívali sme 3D skeny (FLAIR, T1 a T2 modality), ktoré boli postupne rozdelené na individuálne 2D rezy a použité na vstup neurónovej siete s enkodér-dekodér architektúrou. Dataset na trénovanie obsahoval 6720 rezov s rozlíšením 192 x 192 pixelov (po odstránení rezov, ktorých maska neobsahovala žiadnu hodnotu). Využitá loss funkcia bola Combo loss (kombinácia Dice Loss s upravenou Cross-Entropy). Prvá metóda sa zameriavala na využitie predtrénovaných váh z ImageNet datasetu na enkodér U-Net architektúry so zamknutými váhami enkodéra, resp. bez zamknutia a následného porovnania s náhodnou inicializáciou váh. V tomto prípade sme použili len FLAIR modalitu. Transfer learning dokázalo zvýšiť sledovanú metriku z hodnoty približne 0,4 na 0,6. Rozdiel medzi zamknutými a nezamknutými váhami enkodéru sa pohyboval okolo 0,02. Druhá navrhnutá technika používala self-supervised kontext enkodér s Generative Adversarial Networks (GAN) na predtrénovanie váh. Táto sieť využívala všetky tri spomenuté modality aj s prázdnymi rezmi masiek (spolu 23040 obrázkov). Úlohou GAN siete bolo dotvoriť sken mozgu, ktorý bol prekrytý čiernou maskou v tvare šachovnice. Takto naučené váhy boli následne načítané do enkodéru na aplikáciu na náš segmentačný problém. Tento experiment nevykazoval lepšie výsledky, s hodnotou DSC 0,29 a 0,09 (nezamknuté a zamknuté váhy enkodéru). Prudké zníženie metriky mohlo byť spôsobené použitím predtrénovaných váh na vzdialených problémoch (segmentácia a self-supervised kontext enkodér), ako aj zložitosť úlohy kvôli nevyváženému datasetu.
|
|
|
Computer analysis of medical image data
Krajčír, Róbert ; Šmirg, Ondřej (referee) ; Uher, Václav (advisor)
This work deals with medical image analysis, using variety of statisic and numeric methods implemented in Eclipse and Rapidminer environments in Java programming language. Sets of images (slices), which are used here, are the results of magnetic resonance brain examination of several subejcts. Segments in this 3D image are analyzed and some local features are computed, based on which data sets for use in training algorythms are generated. The ability of successful identification of healthy or unhealthy tissues is then practically tested using available data.
|
|
|
Trainable image segmentation using deep neural networks
Majtán, Martin ; Burget, Radim (referee) ; Harár, Pavol (advisor)
Diploma thesis is aimed to trainable image segmentation using deep neural networks. In the paper is explained the principle of digital image processing and image segmentation. In the paper is also explained the principle of artificial neural network, model of artificial neuron, training and activation of artificial neural network. In practical part of the paper is created an algorithm of sliding window to generate sub-images from image from magnetic rezonance. Generated sub-images are used to train, test and validate of the model of neural network. In practical part of the paper si created the model of the artificial neural network, which is used to trainable image segmentation. Model of the neural network is created using the Deeplearning4j library and it is optimized to parallel training using Spark library.
|
|
|
Multiple sclerosis detection
Kopuletý, Michal ; Mangová, Marie (referee) ; Uher, Václav (advisor)
This thesis is focused on detecting multiple sclerosis lesions from magnetic resonance images. Correctly retrieved lesions are very important for medical diagnosis. Detection of lesions using machine learning techniques is quite challenging because of large variability in size, shape and position of lesions in the brain. In the practical part is designed base software, which after completion will classify pixels, so that is possible to find lesions of multiple sclerosis. For classification will be used Support vector machine. Theoretical part describes multiple sclerosis, basic operations performed with biomedical images and data classification.
|
|
|
Work integration options in people with multiple sclerosis: influence of symptoms and other comorbidities
Ulmanová, Alena ; Novotná, Klára (advisor) ; Rodová, Zuzana (referee)
BACHELOR THESIS ABSTRACT Author: Alena Ulmanová Supervisor: Mgr. Klára Novotná Ph.D. Consultant: Mgr. Eliška Rotbartová Title: Work integration options in people with multiple sclerosis: influence of symptoms and other comorbidities Abstract: This theoretical-practical bachelor thesis focuses on the impact of multiple sclerosis symptoms and other comorbidities on the employment of people who are dealing with this diagnosis. The thesis aims to find out what are the main difficulties that limit the ability to work of people with multiple sclerosis, using a questionnaire of work difficulties called Multiple Sclerosis Work Difficulties Questionnaire (MSWDQ-23). The importance of this aim is based primarily on the fact that the disease mainly affects young adults in productive age, for whom employment tends to be an essential part of life and the difficulties associated with reduced work capacity can negatively affect their financial and social situation, but also their general physical condition. The theoretical part of the thesis summarizes the knowledge on the given issue using the current foreign and Czech literature. In the practical part, quantitative research based on a questionnaire survey is used. A standardize questionnaire called MSWDQ- 23 is used, which comprehensively assesses the work difficulties...
|
|
|
Interdisciplinary cooperation in planning care for clients with multiple sclerosis in residential respite service
Linhartová, Petra ; Vrzáček, Petr (advisor) ; Mertl, Jiří (referee)
The diploma thesis deals with the effectiveness of interdisciplinary cooperation by setting up processes focused on communication and information sharing in a multidisciplinary team. It is based on theoretical starting points mainly in connection with a systemic approach to the client. The principle of case management, individual cooperation with the client and the consulting paradigm of social work were applied here. The aim of the work is to identify possible forms of effective interdisciplinary cooperation in the selected facility, using the example of which it presents the elements affecting the quality of outputs of care planning for clients with multiple sclerosis in the conditions of a specific residential respite service. In the theoretical part of the thesis, a unique organization in the Czech Republic, which has been involved in the care of people in a long term, especially in the advanced stage of the disease and which includes a residential respite service, is presented. The mentioned respite service strives to improve the quality of care for clients by setting up effective interdisciplinary cooperation using the Human Centered Design (HCD) method. Through a case study describes the process of introducing change and setting up cooperation processes in a specific multidisciplinary team...
|
|
|
Cognitive rehabilitation of people with multiple sclerosis
Bílková, Tereza ; Novotná, Klára (advisor) ; Krivošíková, Mária (referee)
BACHELOR THESIS ABSTRACT Name, Surname: Tereza Bílková Supervisor: Mgr. Klára Novotná, Ph.D. Consultant: Mgr. Eliška Rotbartová Title: Cognitive rehabilitation of people with multiple sclerosis Abstract: This bachelor thesis focuses on the topic of cognitive rehabilitation in patients with multiple sclerosis, which takes place in a home environment using the computer program HAPPYneuron. Multiple sclerosis is classified as an autoimmune disease, during which the cells of our central nervous system are damaged. The aim of this bachelor's thesis is to test whether cognitive rehabilitation of 16 weeks, conducted through the computer program HAPPYneuron, has a positive impact on spatial orientation, short-term memory, attention and speech in MS patients. In the theoretical part of the bachelor thesis I describe the actual issue of multiple sclerosis, including the clinical picture and types of multiple sclerosis. I then go on to describe in more detail the most affected domains of cognitive function, their examination and the possibilities of cognitive rehabilitation. Last but not least, I describe cognitive rehabilitation using computer programs, with a more detailed description of the HAPPYneuron program. In the practical part, three patients were first examined by a clinical neuropsychologist. Based on this,...
|
|
|
Multiple sclerosis - clinical and paraclinical markers for monitoring disease activity and factors influencing its course
Šťastná, Dominika ; Horáková, Dana (advisor) ; Libertínová, Jana (referee) ; Taláb, Radomír (referee)
Multiple sclerosis (MS) is a chronic neurological disease that, without treatment, leads over years to decades to severe disability in most patients. We cannot cure the disease, but there is growing evidence that early initiation of anti-inflammatory therapy and management of associated comorbidities has a major impact on its course. Patient registries have an irreplaceable contribution to evaluating factors influencing the MS course and the monitoring of therapeutic agents in real clinical practice. First, this thesis evaluated therapy management trends between 2013 and 2021 based on data from the Czech National MS Registry (ReMuS). Subsequently, the paper responds to the onset of the covid-19 pandemic through registry data and addresses this issue in the context of MS. The proportion of patients in ReMuS treated with high-efficacy disease-modifying therapies (HE-DMT) increased from 16.2% to 37.1% between 2013 and 2021, and the proportion of treatment-naive patients initiating HE-DMT increased from 2.1% to 18.5%. Regarding covid-19 infection, we determined that higher body mass index, older age, recent high-dose glucocorticoid treatment, and anti-CD20 therapy were independent variables associated with pneumonia based on data from 958 MS patients with a history of covid-19. Further, we analyzed...
|
|
|
Eye-Tracking Control of an Adjustable Bed
Kopeček, Martin ; Kremláček, Jan (advisor) ; Čapek, Lukáš (referee) ; Komzák, Martin (referee)
The origin of this work was based on the need to control an electric positioning bed by patients with no or significantly reduced upper limb motor skills. The key point and objective of the dissertation study was to develop non-contact alternatives to manual controls and to verify that the eye-tracking technique is usable and offers patients a new level of increased self-sufficiency. The thesis is organized into three related parts with experiments conducted at the detached departments and in the laboratory. After an introductory section covering the stages of development and current progressive trends in eye movement tracking, an experimental study of the applicability of bed control with the role of alternating head and leg position changes using on-screen graphical controls is described. This stage was conducted using a virtual bed. In a group of 17 patients with diagnoses of a pentaplegia, tetraplegia, high paraplegia, myopathy, and spinal muscular atrophy, the overall time to solve the task was 67.1 s (median) with a large interindividual variability with interquartile range from 56.7 s to 92.9 s. The solution efficiency (100 % matched to optimal performance) was 45.5 (34.9; 62.0) %. Within each group patients achieved different results for both studied parameters. When evaluating the features of the...
|
guest ::
