Comprehensive feature extraction for objective Dynamic Gait Index assessment of risk of falls in the elderly

• Autores: Miralles, Mónica Teresita; Ghersi, Ignacio; Vecchio, R.; Paterson, R.; Pérez Akly, Manuel; Ferrando, María H.; Paterson, A; Alvarez, F.
• Año de publicación: 2013
• Idioma: inglés
• Tipo de recurso: artículo
• Fuente: Journal of Physics: Conference Series. Vol.477, 2013
• Materia: ADULTOS MAYORES; ACCIDENTES POR CAIDAS; FACTORES DE RIESGO; EQUILIBRIO; PREVENCION Y CONTROL
• Institución: Pontificia Universidad Católica Argentina

Fil: Miralles, Mónica Teresita. Pontificia Universidad Católica Argentina. Facultad de Ciencias Fisicomatemáticas e Ingeniería. Laboratorio de Biomecánica e Ingeniería para la Salud; Argentina
Fil: Miralles, Mónica Teresita. Universidad de Buenos Aires. Facultad de Arquitectura, Diseño y Urbanismo. Centro de Investigación en Diseño Industrial; Argentina
Fil: Ghersi, Ignacio. Pontificia Universidad Católica Argentina. Facultad de Ciencias Fisicomatemáticas e Ingeniería. Laboratorio de Biomecánica e Ingeniería para la Salud; Argentina
Fil: Vecchio, R. Pontificia Universidad Católica Argentina. Facultad de Ciencias Fisicomatemáticas e Ingeniería. Laboratorio de Biomecánica e Ingeniería para la Salud; Argentina
Fil: Paterson, R. Universidad de Buenos Aires. Facultad de Arquitectura, Diseño y Urbanismo. Centro de Investigación en Diseño Industrial; Argentina
Fil: Pérez Akly, Manuel. Unidad Asistencial Dr. César Milstein. Servicio de Neurología; Argentina
Fil: Ferrando, María H. Unidad Asistencial Dr. César Milstein. Servicio de Neurología; Argentina
Fil: Paterson, A. Universidad de Buenos Aires. Facultad de Arquitectura, Diseño y Urbanismo. Centro de Investigación en Diseño Industrial; Argentina
Fil: Alvarez, F. Unidad Asistencial Dr. César Milstein. Servicio de Neurología; Argentina
Abstract. The evaluation and understanding of human balance and risk of falls, particularly in the elderly, is a growing field of study. Complementeing classic clinical tests (relying on comprehensive, although mostly qualitative assessments) with new devices that acquire precise patterns associated to the balance process, appears as a great oportunity in this area. Objective features can be obtained from these patterns, summarizing relevant data about balance and gait processes. This work presents results from the Dynamic Gait Index assessment over 42 subjects, while wearing a specially designed 3D-linear acceleration sensing device. With knowledge about the test, features related to step frequency, energy expenditure and changes in signal shape were defined and obtained from the three acceleration signals. The evolution of these features over time is of interest, so a moving-window strategy was implemented. Signal processing strategies were tested and improved, and results were correlated with the clinical scores given by the physicians to better understand them, initially test their reliability and select the most suitable ones. Results showed promise for this strategy in providing uniform, meaningful and new interpretations about different processes involved in the stages of the test, opening the possibility to develop custom user interfaces for clinical use.