Eye-Dialect System for Aphasia Person Computing K-Means and EM Clustering Algorithms

Locked in syndrome (LIS) person is deprived to communicate in conventional way. Though suffering from diseases like spinal cord injury, amyotrophic lateral sclerosis (ALS), cerebral palsy, etc., person retains the ability to control their eye movements. This ability of voluntary eye movements for human computer interface (HCI) can be used as alternative orientation. In this paper, for different eye gestures acquired using Ag-AgCl electrodes and signal conditioning board AD8232 and interfacing with arduino AT mega 2560 output messages obtained by K-means clustering algorithm. Output for different gestures like voluntary blink, involuntary blink, gaze, vertical upward movement, vertical downward movement, horizontal left to right, horizontal right to left, diagonal left to right, diagonal right to left, etc. are used to communicate different messages like ?need water?, ?need food?, ?feeling uneasy?, etc. using K-means clustering algorithm, the LIS person thus can communicate with outer world, enriching the quality of life. Efficiency of 98% is obtained for voluntary and involuntary blink respectively and 100% for gaze gesture.