Hledat
Zobrazují se záznamy 1-1 z 1
Metody a parametry adaptivní segmentace EEG záznamů., Methods and parameters for adaptive segmentation of EEG recordings
; Vedoucí práce: Krajča Vladimír; Oponent práce: Lhotská Lenka (České vysoké učení technické v Praze. Vypočetní a informační centrum.Czech Technical University in Prague. Computing and Information Centre., 2016-05-20)
nestacionarity EEG) je proces adaptivní segmentace (AS) signálu, jehož kvalita se významně podílí na kvalitě celého systému. Právě adaptivní segmentace EEG záznamů a vliv různých vstupních parametrů na výsledek segmentace je předmětem této práce. Byla vytvořena...
Visual review of long-term EEG recordings, as performed in common medical practice, is strongly time-consuming. Therefore, there is an effort to develop the semi-automatic and automatic systems, which helps physicians to make their work faster and easier. Adaptive segmentation (AS) is an essential first step of such systems and its precision has significant influence to performance of whole system. AS of EEG records and influence of its input parameters to segmentation quality is subject of this diploma thesis. MATLAB software and methodology for comparing of various AS methods was developed in this thesis and Varri?s method with two connected windows was selected as most suitable for this. The effect of various input parameters and its values on artificial and real EEG signals segmentation using this method was then tested. Sinusoidal and autoregressive artificial signals and normal, epileptic and neonatal EEG recordings was used for testing. Optimal values of parameters for segmentation of clinical EEG records was specified. The software for methods comparison and for input parameters testing is the practical outcome of this thesis. A detailed study of segmentation performance according to the input parameters is the theoretical outcome of this thesis. Also the ranges of parameters values for optimal segmentation of clinical EEG recordings were established....
Visual review of long-term EEG recordings, as performed in common medical practice, is strongly time-consuming. Therefore, there is an effort to develop the semi-automatic and automatic systems, which helps physicians to make their work faster and easier. Adaptive segmentation (AS) is an essential first step of such systems and its precision has significant influence to performance of whole system. AS of EEG records and influence of its input parameters to segmentation quality is subject of this diploma thesis. MATLAB software and methodology for comparing of various AS methods was developed in this thesis and Varri?s method with two connected windows was selected as most suitable for this. The effect of various input parameters and its values on artificial and real EEG signals segmentation using this method was then tested. Sinusoidal and autoregressive artificial signals and normal, epileptic and neonatal EEG recordings was used for testing. Optimal values of parameters for segmentation of clinical EEG records was specified. The software for methods comparison and for input parameters testing is the practical outcome of this thesis. A detailed study of segmentation performance according to the input parameters is the theoretical outcome of this thesis. Also the ranges of parameters values for optimal segmentation of clinical EEG recordings were established....