ABSTRACT
People with severe upper limb paralysis use devices that monitor head movements to control computer cursors. The three most common methods for producing mouse button clicks are dwell-time, sip-and-puff control, and voice-recognition. Here, we tested a new method in which small tooth-clicks were detected by an accelerometer contacting the side of the head. The resulting signals were paired with head tracking technology to provide combined cursor and button control. This system was compared with sip-and-puff control and dwell-time selection. A group of 17 people with disabilities and ten people without disabilities tested each system by producing mouse clicks as inputs to two software programs. Tooth-click/head-mouse control was much faster than dwell-time control and not quite as fast as sip-and-puff control, but it was more reliable and less cumbersome than the latter.
