Tongue Movement Estimation Based on Suprahyoid Muscle Activity

1. Introduction

To support independent living of people with severe quadriplegia caused by cervical cord injury or muscular dystrophy, it is important to understand one’s intention from biological signals that can be measured noninvasively. Electroencephalography (Wolpaw et al., 2002; Scherer et al., 2004), head movement (Harwin et al., 1990; Nguyen et al., 2004), jaw movement (Jacobs et al., 1997; Niikawa et al., 2006), eye movement (LaCourse et al., 1990; Duchowski, 2002), voice (Clark et al., 1977; Simpson et al., 2002), and breathing have all been used for the communication of intentions by people with severe disabilities. Each intention-communication method entails benefits and shortcomings. For example, although the social expectations for implementing electroencephalography are high, its use requires high levels of concentration over long periods in real-life environments, where a person receives various external stimuli. Although voluntary movements of the head and jaw can readily express one's intentions, they might not be applicable to a person with cervical cord injury because of insufficient cervical stability. Furthermore, although intention communication by eye movement or voice is effective under environments in which external disturbances are removed or controlled to the greatest extent possible, precise isolation and extraction of these signals is difficult when, for example, one is driving an electric wheelchair in an urban setting. Examining many methods of intention communication including those described above, we must develop them and improve their precision to prepare several alternatives that are readily applicable and which are convenient for individuals with various disabilities.

Against this background, intentional tongue movement is attracting attention in recent years. The motor function of the tongue often remains intact even in cases of severe movement paralysis. Therefore, tongue movements offer great potential for the design of novel highly efficient human-machine interfaces for alternative communication and control. Numerous approaches for deriving control signals from intentional tongue motions have been proposed: detection of the tongue position via measurement of magnetic fields of a permanent magnet attached to the tongue (Sonoda, 1974; Huo et al., 2008), detection of lingual proximity by light-emitting diodes and photodiodes placed on an artificial palate plate (Wrench et al., 1998; Saponas et al., 2009), measurement of the tongue force applied to a force sensor array mounted on an artificial palate plate (Ichinose et al., 2003; Terashima et al., 2010), and direct tongue manipulation of a joystick or switch inserted into the oral cavity (Niikawa et al., 2006). However, such methods require the insertion of a measuring instrument into the oral cavity, which entails certain risks and discomfort to the patient such as increased psychological stress, oral health problems, obstruction of speaking and drinking, battery fluid leakage, electric shock, and suffocation by accidental ingestion. Therefore, we have been developing a novel intention-communication method based on surface electromyography (EMG) signals (Sasaki et al., 2012a, 2012b, 2013a, 2013b).