Caracterización de la parálisis cerebral mediante la estimación del movimiento ocular
Cargando...
Archivos
Autores
Gonzáles Huertas, Jully Pauline
Director
Tipo de contenido
Document language:
Español
Fecha
Título de la revista
ISSN de la revista
Título del volumen
Documentos PDF
Resumen
La Parálisis Cerebral (PC) se considera un problema de salud pública y la causa más común de discapacidad motora en niños, con una incidencia de 2 a 2.5 casos por 1,000 nacidos vivos. Entre el 50-90 % de los pacientes con esta condición presentan problemas visuales, por lo que esta población exhibe anomalías visuales sensoriales y motoras a tasas que superan las detectadas en niños neurológicamente normales. Este trabajo propone una caracterización automática de la parálisis cerebral mediante la estimación del movimiento ocular durante una tarea de búsqueda de la mirada. Para esto, se realiza una estimación del movimiento del ojo utilizando un enfoque del flujo óptico, en un conjunto de datos organizado en dos grupos: niños con parálisis cerebral y controles. A partir del campo de movimiento obtenido, se lleva a cabo una caracterización cuantitativa del movimiento del ojo a través de histogramas de orientación y magnitud. El resultado temporal se acumula en un histograma y una distancia de correlación entre histogramas cuantifica las diferencias entre grupos. El enfoque propuesto fue evaluado con un grupo experimental de 16 pacientes con parálisis cerebral y 16 sujetos controles. Al evaluar con la prueba estadística de Wilcoxon, los resultados muestran diferencias significativas entre los pacientes con parálisis cerebral y los controles, obteniendo un valor de p0.01 para las distancias de orientación en ambas tareas visuales y un valor de p0.05 para las distancias de magnitud en la tarea de búsqueda suave. Haciendo de esta estrategia una forma de caracterizar los patrones del movimiento ocular en los niños con parálisis cerebral.
Abstract: Cerebral Palsy (CP) is considered as a public health problem and the most common cause of motor illness in children with an incidence of 2 to 2.5 cases per 1,000 live births. Between 50-90 % of the patients with this condition present visual problems, reason why this population exhibits visual sensory and motor abnormalities at rates exceeding those detected in neurologically normal children. This work proposes an automatic characterization of cerebral palsy by estimating the ocular motion during a gaze pursuit task. For this, an estimation of the eye movement is performed using an optical flow approach, at a dataset organized in two groups: cerebral palsy children and controls. From the motion field obtained, a quantitative characterization of the eye movement is carried out through velocity histograms of magnitude and orientation. The temporal result is accumulated in a histogram and a correlation distance between histograms quantifies the differences between groups. The proposed approach was evaluated with an experimental group of 16 cerebral palsy patients and 16 control subjects. When evaluating with the Wilcoxon statistical test, results show significant differences between cerebral palsy patients and controls, obtaining a p-value 0.01 for the orientation distances in both visual tasks and a p-value 0.05 for the magnitude distances in the smooth pursuit task. Making this strategy a way to characterize ocular movement patterns in children with cerebral palsy.
Abstract: Cerebral Palsy (CP) is considered as a public health problem and the most common cause of motor illness in children with an incidence of 2 to 2.5 cases per 1,000 live births. Between 50-90 % of the patients with this condition present visual problems, reason why this population exhibits visual sensory and motor abnormalities at rates exceeding those detected in neurologically normal children. This work proposes an automatic characterization of cerebral palsy by estimating the ocular motion during a gaze pursuit task. For this, an estimation of the eye movement is performed using an optical flow approach, at a dataset organized in two groups: cerebral palsy children and controls. From the motion field obtained, a quantitative characterization of the eye movement is carried out through velocity histograms of magnitude and orientation. The temporal result is accumulated in a histogram and a correlation distance between histograms quantifies the differences between groups. The proposed approach was evaluated with an experimental group of 16 cerebral palsy patients and 16 control subjects. When evaluating with the Wilcoxon statistical test, results show significant differences between cerebral palsy patients and controls, obtaining a p-value 0.01 for the orientation distances in both visual tasks and a p-value 0.05 for the magnitude distances in the smooth pursuit task. Making this strategy a way to characterize ocular movement patterns in children with cerebral palsy.