Algorithms for target prediction for computer users with athetosis

Abstract

Athetosis is a movement disorder that afflicts numerous persons with cerebral palsy, resulting in significant problems in their control of computer interfaces. As a step toward increasing the efficiency of icon selection by computer users with athetosis, we have implemented three techniques to reduce the time of target acquisition: transition assistance via directional gain variation based on target prediction during initial movement toward the target, settling assistance via gain reduction when in the vicinity of a predicted target, and expansion of the predicted target as the cursor approaches it. The paper describes each method, and presents results from evaluation of each method using a closed-loop model of a human subject with athetosis, trained using recorded data, at three different severity levels.