Severely disabled persons with impairments of the nervous system, especially from brain and spinal cord injuries, find it difficult to carry out daily tasks without assistance. The Christopher and Dana Reeve Foundation recently reported that nearly 1 in 50 people in the United States are living with paralysis, and nearly one-fifth of these individuals require continuous help. Every year 11,000 new cases occur as a result of car accidents, acts of violence, and falls. More than one-half of these victims are between the ages of 16 and 30 and will require special care for the rest of their lives.
Loss of motor control significantly reduces quality of life; individuals cannot perform many of the daily tasks that able-bodied people take for granted such as turning on a light or opening a door. It also forces individuals to be continuously dependent upon a family member or dedicated caregiver, which can significantly increase their healthcare costs. Wheeled mobility for transportation inside and outside the home is a key enabling factor for those suffering from paralysis. Most use electrically powered wheelchairs (PWC) that allow them to perform daily tasks with greater independence. However, most PWCs are controlled by a joystick, which requires hand function often lacking in people with severe disabilities.
The Tongue Wins
Dr. Maysam Ghovanloo, Assistant Professor and Director of the GT-Bionics Laboratory at Georgia Institute of Technology, and his research team have developed a tongue-operated AT called the Tongue Drive System (TDS) that is unobtrusive, wearable, and wireless and could substitute for many arm and hand functions. “Tongue Drive is designed to enable individuals with severe spinal cord injuries who are completely paralyzed. The goal is for users to be able to drive their wheelchairs, operate their computers, and generally control their environment in a more independent fashion,” says Ghovanloo. The TDS only requires that a user be able to move his or her tongue, a function that is usually maintained even in severe spinal cord injuries. “Even a person who cannot breathe on his own or talk can still move his tongue,” adds Ghovanloo.
A tongue-driven device offers several other distinct advantages. The area in the brain that controls tongue and mouth movements is nearly the same size as the area that controls the fingers and hand. Therefore, the tongue is capable of sophisticated tasks and precise motor control. The tongue is fast and agile, and the tongue muscle, similar to the heart muscle, is resistant to fatigue as evidenced by everyday talking and eating. Finally, because the tongue is hidden in the mouth, it provides TDS users a certain degree of privacy that is not available from devices that rely on head control or eye movements.
It’s On the Tip of My Tongue
Tongue movements change the magnetic field around the mouth, which is detected by small magnetic sensors mounted on the headset near the user’s cheeks. “The magnetic sensors are the same type used in GPS devices to sense the earth’s magnetic field, so they are very sensitive,” says Ghovanloo. A wireless transmitter in the headset sends information from the sensors to the computer that translates the changes in magnetic field into the desired user commands. Depending on the computer processing speed, the time between tongue movement and action occurs within milliseconds, a delay that is not perceived by the user.
What Your Tongue Can Do for You
Before using the TDS, the user must train the system to link each designated tongue movement to an action. Training requires commands to be repeated several times for computer recognition, but is relatively fast. Each command is defined by collecting changes in the magnetic field in relation to the tongue position. “The device currently allows for six commands that are all available to the user at the same time. The ultimate goal is to allow the user to designate every tooth to a command,” says Ghovanloo. Most adult humans have 32 teeth, and several commands can be linked to a combination of teeth or tongue gestures, making the possibilities countless.
Through a wireless local area network (WLAN), the TDS can be linked to currently available SmarthomeTM technologies that control household appliances, lights, locks, heating/air conditioning, etc. The device also could control prosthetic arms or legs. “A prosthetic limb can have all the capabilities of a natural limb, but if you cannot control it, those capabilities are useless,” says Ghovanloo. “With the Tongue Drive System, the user can take advantage of all of the capabilities of the tongue to control his or her prosthesis,” he adds.
Although there can be some degree of error when translating movement into action, the TDS has shown a high degree of accuracy. Most errors result from user mistakes, such as making the incorrect movement. In initial experiments, all of the subjects tested were naïve to the system and sometimes forgot the correct tongue movement-action association. Subjects with more system experience had a very low error rate, and accuracy improved with time. Dr. Ghovanloo is conducting long-term studies to evaluate this technology and speed of learning in individuals with high-level disabilities. However, for the most part, preliminary learning is very fast and easy to remember, and is expected to become intuitive with time.
Some tongue movements – such as speech, swallowing, coughing, or sneezing – could interfere with proper functioning of the TDS if the tongue commands are not defined properly. Since those natural tongue movements often take place in the sagittal plane (center of the mouth), TDS commands are defined using tongue motions on either the left or right side of the mouth. Eating is one exception, however, because the tongue moves everywhere in the mouth. To address this type of interference, the device has a predefined stand-by or sleep command. When ready to eat, the user issues the sleep command by touching the tongue to the left cheek and holding it there for 3 seconds. After eating, the same command brings TDS back to active mode. TDS flexibility in defining a unique set of tongue commands for every user is especially useful for some stroke patients who may lose mobility on one side of their body. These individuals can define all their tongue commands on the functional side of their mouth.
Dr. Ghovanloo’s device has gone through six or seven rounds of small-scale experimental trials in able-bodied subjects, with improvements made in every successive round. The first clinical trial was conducted in 2009 at the Shepherd Center in Atlanta, one of the nation’s top rehabilitation centers. Thirteen subjects with high-level spinal injuries participated in this trial and the results were published in the Journal of Neural Engineering. Dr. Ghovanloo has secured additional funding through NIBIB to conduct a multicenter trial in which the magnet will be attached to the tongue by piercing rather than glue. The TDS tongue piercing is much like an ornamental tongue piercing except that, instead of jewelry, a tiny magnet is embedded in the stud. This allows for long-term, semi-permanent use of the device, unlike the previous method that only lasts for a few hours before the magnet must be reattached. The magnet has been completely enclosed inside a laser-welded titanium sphere that costs about $50.
Although the current TDS is state of the art, Dr. Ghovanloo already has plans for improvements. Eventually, a dental implant could replace the headset that transmits the magnetic field changes, an option for users who want their AT completely hidden from sight. To further streamline the device, it will be linked to a wearable smartphone, such as an iPhoneTM, that transmits the signals from the magnet to the user’s network (e.g., house, wheelchair) eliminating the need to carry a computer on a wheelchair.
Hold Your Tongue
Dr. Ghovanloo’s tongue device has already received significant attention and media coverage. “There is a huge interest in this technology because there is nothing on the market that can provide this level of capability,” says Ghovanloo. The device is expected to be ready for the commercial market by the end of 2011.
This work is supported in part by the National Science Foundation, the Dana and Christopher Reeve Foundation, the U.S. Army Research Office, and American Recovery and Reinvestment Act funds through the National Institute of Biomedical Imaging and Bioengineering.