Description
Analyzing human pupillary behavior is a non-invasive method for assessing neurological activity. Changes in this behavior correlate to various health conditions, such as: Parkinson's, Alzheimer's, autism and diabetes. In comparison to other methods of neurological evaluation, analyzing pupillary behavior is a simple and low-cost method that can be used as a complementary diagnosis. Therefore, the goal of this work is to develop a system to evaluate pupillary behavior in an automated way. The system consists of a portable recording device, called a pupilometer, and a software based on computer vision. The system is capable of stimulating, recording, measuring and extracting important features of the human pupillary behavior. The results demonstrate that the system is a fast and accurate solution, with 97.25% of accuracy in the segmentation and pupil measurement on the images.
Recommended Citation
Silva, Cleyton Rafael Gomes; Gonçalves, Cristhiane; Camilo, Eduardo; Laureano, Gustavo; Siqueira, Joyce; Boaretti, Fábio; and da Costa, Ronaldo Martins, "Sistema Automatizado para Avaliação do Comportamento Pupilar Humano baseado em Visão Computacional" (2017). AMCIS 2017 Proceedings. 5.
https://aisel.aisnet.org/amcis2017/Spanish/Presentations/5
Sistema Automatizado para Avaliação do Comportamento Pupilar Humano baseado em Visão Computacional
Analyzing human pupillary behavior is a non-invasive method for assessing neurological activity. Changes in this behavior correlate to various health conditions, such as: Parkinson's, Alzheimer's, autism and diabetes. In comparison to other methods of neurological evaluation, analyzing pupillary behavior is a simple and low-cost method that can be used as a complementary diagnosis. Therefore, the goal of this work is to develop a system to evaluate pupillary behavior in an automated way. The system consists of a portable recording device, called a pupilometer, and a software based on computer vision. The system is capable of stimulating, recording, measuring and extracting important features of the human pupillary behavior. The results demonstrate that the system is a fast and accurate solution, with 97.25% of accuracy in the segmentation and pupil measurement on the images.