Título: An Intelligent Human-Machine Interface Based on Eye Tracking to Afford Written Communication of Locked-In Syndrome Patients
Autores: Leonel, Amanda; Lima Neto, Fernando B. de; Oliveira, Sérgio Campello; Filho, Hugo Serrano B.
Resumo: Patients with ‘Locked-in Syndrome’ (LIS) report rare and severe clinical cases as they become paralyzed and voiceless. Due to a brain damage, patients lose movements of their voluntary muscles and also,their ability to speak. In spite of that, their cognitive functions remain in perfect order. In general, the only voluntary movement that still remains is the motor control of the eyes and eyelids. This work presents a proof-of-concept,including an algorithm, based on an Intelligent Human-Computer Interface to enable written communication of paralytic patients. The proposed system uses techniques from Computer Vision and Artificial Intelligence in order to acquire and detect eye-eyelid movements affording written communications of LIS patients. The processed images result in low resolution patterns of the patients’ eye-eyelid. Artificial Neural Networks are then trained in order to recognize such patterns and relate them to written symbols that are presented in a computer screen. The high rate of successand performance on the system, obtained in experiments demonstrate the feasibility of the proposed interface model.
Palavras-chave: Human-Computer Interface; Paralysis; Locked-in Syndrome; Eye-tracking; Artificial Neural Networks
Páginas: 7
Código DOI: 10.21528/lmln-vol9-no4-art3
Artigo em PDF: vol9-no4-art3.pdf
Arquivo BibTex: vol9-no4-art3.bib