Catálogo de publicaciones - tesis
Título de Acceso Abierto
Algoritmos avanzados para la detección del síndrome de apnea-hipopnea obstructiva del sueño
Román Emanuel Rolón Hugo Leonardo Rufiner Hugo Aimar Gastón Schlotthauer Marcelo Risk Juan Carlos Gómez Rubén Daniel Spies
acceptedVersion.
Resumen/Descripción – provisto por el repositorio digital
Obstructive sleep apnea-hypopnea syndrome is one of the most common sleep disorders in the general population. It is estimated that this condition affects between 3% and 5% of the adult population worldwide and increases with age. Abnormal respiratory events occur as a consequence of an anatomical-functional alteration of the upper airway producing its narrowing (hypopnea) or blockage (apnea). The severity of the syndrome is quantified by the apnea-hypopnea index, which represents the rate of events per hour of sleep. This thesis addresses the design, development, implementation and evaluation of three methods to support the diagnosis of the pathology by automatically recognizing apnea and hypopnea events using only blood oxygen saturation signals. In a first stage, two methods for feature selection called MDAS and MDCS based on sparse representations of signals in terms of discrete dictionaries are presented . Simultaneously, a new binary discriminative measure denoted by DCAF, which is able to efficiently quantify the degree of discrimination of atoms in a given dictionary, is introduced. Results show that MDCS significantly outperforms the other three state of the art methods. In addition, a new iterative method called DAS-KSVD for the learning of structured dictionaries in the context of multi-class classification problems obtaining promising results is presented.Palabras clave – provistas por el repositorio digital
Apnea syndrome; Pulse oximetry; Signal processing; Sparse representations; Pattern recognition; Machine learning; Síndrome de apnea; Oximetría de pulso; Procesamiento de señales; Representaciones ralas; Reconocimiento de patrones; Aprendizaje maquinal
Disponibilidad
| Institución detectada | Año de publicación | Navegá | Descargá | Solicitá |
|---|---|---|---|---|
| No requiere | 2019 | Biblioteca Virtual de la Universidad Nacional del Litoral (SNRD) |
|
Información
Tipo de recurso:
tesis
Idiomas de la publicación
- español castellano
País de edición
Argentina
Fecha de publicación
2019-07-04
