A probabilistic algorithm for estimating the intention of computer users with movement disorders

Standard computer interfaces are inefficient for many users who have movement disorders. We believe this stems from the standard view that interface devices accurately represent the intention of the user. For those who have movement disorders, such as dystonia, it may be more effective to view the computer interface as an instrument used to estimate the user´s intent. We develop a probabilistic algorithm to design an estimation device for computer input. We also implement the algorithm as part of text-entry software and perform a preliminary case study on a patient with dystonia, using a joystick as a text input device.