- author = "Ling Guo",
- title = "Soft Computing Techniques for Advanced Epileptic {EEG} Analysis and Classification",
- school = "Facultade de Informatica, Universidade da Coruna",
- year = "2011",
- address = "Spain",
- month = "26 " # may,
- keywords = "genetic algorithms, genetic programming, inteligencia artificial",
-
URL = "
https://dialnet.unirioja.es/servlet/tesis?codigo=43887",
- broken = "http://www.fic.udc.es/NewsContent.do?newsId=25045&urlCurrent=ViewPaginatedCategory.do",
- broken = "http://tesis.com.es/documentos/soft-computing-techniques-for-advanced-epileptic-eeg-analysis-and-classification/",
- size = "210 pages",
-
abstract = "Epilepsy is an abnormal neurological status that makes
people susceptible to brief electrical disturbance in
the brain thus producing a change in sensation,
awareness, and/or behaviour and is characterised by
recurrent seizures. It affects up to 1percent of the
population in the world. Two-thirds of the epileptic
patients can be treated through medications. Another
8percent may benefit from surgery. But 25percent of
people with epilepsy continue to have seizures and no
treatment suits them.
Electroencephalogram (EEG) is the recording of electrical activity of the brain and it contains much valuable information for understanding epilepsy. In clinic environments, the neurologists have to continuously observe the EEG recordings for better understanding epilepsy, which is time-consuming and tedious. Thus, efforts on developing automatic epileptic seizure detection on EEG background are of great importance for epilepsy diagnosis and treatment, and to improve the clinical assistance and, at last, for enhancing the whole health system.
This research successfully combines soft computing techniques of Artificial Neural Networks (ANNs) and Genetic Programming (GP) with signal processing tools of wavelet transform and multiwavelet analysis for advanced epileptic EEG signal analysis and classification. The main objectives of this dissertation are specifically:
on scalar wavelet processing technique. Scalar wavelets are efficient in non-stationary signal analysis. Amounts of classical features based on wavelet analysis have been used for EEG classification by many researches. In this study, new features as Relative Wavelet Energy (RWE) and Line Length (LL) are introduced and extracted from wavelet decomposed EEG signals. Combing these new extracted features with ANNs aims to distinguish epileptic and nonepileptic EEG recordings.",
-
abstract = "Proposing a novel detection method based on
multiwavelet processing technique. Multi-wavelets are
parts of wavelet theory, however, they have some
difference comparing with scalar wavelets. It is usual
to apply scalar wavelets to analyse EEG signals, while,
using multiwavelets to process EEG signals is an
untapped research filed. In this proposed method,
multiwavelet transform analyses EEG signals through
decomposing the signal into several narrow frequency
bands. Then, features sensitive in detecting epileptic
activity are extracted from the decomposed signals.
Finally, the seizure detection procedure is completed
through inputting the extracted feature space into a
classifier system. Additionally, comparisons between
multiwavelets and scalar wavelets on EEG signal
analysis are also investigated in this method.
Applying GP to perform automatic feature extraction. The purpose of this study is to improve the performance of KNN classifier and reduce the input feature dimensionality simultaneously. GP is used to create GP-based features from a set of classical features for detecting epileptic seizures. GP is an automated routine in the family of Evolutionary Computation (EC) that can be used to generate optimal, artificial features. The features are optimal in the sense that the heuristics of EC maximise an objective function, which measures the performance of the artificial features in distinguishing epileptic activity from non-epileptic activity. The obtained GP-based features are thought to be artificial because GP returns computer-crafted results that might not have physical meanings.
Over the years, it has become increasingly clear that the areas of neuroscience,soft computing techniques and EEG signal analysis are not mutually exclusive research areas. Rather, they represent different aspects and any new knowledge gained from one may be a stepping stone for the others. Hopefully, the results of this research can lead to a better diagnosis and treatment of epileptic seizures and an improved quality of life for the millions of persons affected by epilepsy.",
-
abstract = "INTRODUCCION
La epilepsia es un estado neurologico anormal provocado por una perturbacion electrica anomala y breve en una zona del cerebro, lo que produce un cambio en la sensacion, la conciencia y el comportamiento, y se caracteriza por convulsiones recurrentes. Afecta al 1percent de la poblacion en todo el mundo. Dos tercios de los pacientes epilepticos pueden ser tratados con medicamentos, mientras que otro 8percent se pueden beneficiar de la cirugia. Sin embargo, el 25percent de las personas con epilepsia seguiran teniendo convulsiones y no podran ser tratadas.
El Electroencefalograma (EEG) es el registro de la actividad electrica del cerebro y contiene mucha informacion valiosa para la comprension de esta enfermedad. En los entornos clinicos, los neurologos han de observar continuamente el EEG para comprender mejor la epilepsia, proceso que es largo y tedioso. Por lo tanto, los esfuerzos para el desarrollo de sistemas de deteccion automatica de ataques epilepticos mediante el analisis de las senales de EEG son de gran importancia para el diagnostico de la epilepsia y su tratamiento.
Esta investigacion combina tecnicas de Soft Computing como Redes Neuronales Artificiales (RR.NN.AA.) y Programacion Genetica (PG) con herramientas de procesamiento de senal, transformada wavelet y multiwavelet, para realizar un analisis avanzado y clasificacion de la senal de EEG relacionada con la enfermedad de epilepsia.
Concretamente, los principales objetivos de esta Tesis son:
Desarrollar un modelo para la deteccion de crisis epilepticas a traves de la extraccion de nuevas caracteristicas basadas en el analisis escalar wavelet. En este estudio, este analisis se utiliza para extraer nuevas caracteristicas con el objetivo de clasificar las senales de EEG. Estas nuevas caracteristicas se combinan con RR.NN.AA. con el objetivo de distinguir entre grabaciones de EEG epilepticos y no epilepticos.
Proponer un nuevo metodo de deteccion basada en la tecnica de procesamiento multiwavelet. Esta es parte de la teoria de wavelets, sin embargo, tienen alguna diferencia en comparacion con los wavelets escalares. La aplicacion de wavelets escalares para analizar las senales de EEG es una tarea habitual, mientras que el uso de multiwavelets para procesar las senales de EEG es un campo de investigacion apenas sin explotar.
Aplicar PG para realizar la extraccion automatica de caracteristicas. El proposito de este estudio es mejorar los resultados ofrecidos por algoritmo de clasificacion del vecino mas cercano (K-Nearest Neighbor, KNN) y reducir la dimensionalidad del conjunto de entrada simultaneamente.",
-
abstract = "METODOLOGIA
Para conseguir estos objetivos, en este trabajo se propone el uso de una metodologia que permita aplicar dichas tecnicas de la siguiente manera:
En el primer metodo, las senales de EEG en primer lugar se descomponen mediante la transformada wavelet BibTeX entry too long. Truncated
