Individuals suffering from speech disorders or severe muscular weakness soon may have a way to communicate with doctors and loved ones just by breathing their words. A new augmentative and alternative communication (AAC) device uses an algorithm that learns speech patterns from a patient’s breath, and researchers say the technology could prove particularly useful in intensive care units by establishing early diagnosis of locked-in syndrome.
Children and adults who are unable to communicate through regular speech must make use of body language, gestures, or AAC tools to express themselves effectively. According to the American Speech-Language-Hearing Association (ASHA), AAC methods range from a pen and paper to electronic devices that generate speech.
Speech-generating devices (SGDs), famously used by Stephen Hawking and the late Roger Ebert, provide users with pages of pictures, symbols, and letters that produce digitized speech when selected. Output from these devices typically is quite slow and is dependent upon user skill.
Scientists at Loughborough University in the U.K. have proposed a new SGD system that cuts out the pictures and allows patients to communicate their words directly and with very little physical effort. While previous SGD systems required some sort of physical cue — blinking or sniffing, for example — to select a picture, the Loughborough team has developed a system based on breath.
“What we are proposing is a system that learns with the user to form an effective vocabulary that suits the person rather than the machine,” said Dr. David Kerr, senior lecturer at Loughborough, to Leicester Mercury.
In a YouTube video uploaded by the University, scientists demonstrated how the system works. By analyzing and learning the breath patterns of individual patients, the machine can develop an algorithm that then translates these breaths into digitally communicated words.
Kaddour Bouazza-Marouf, a professor of mechatronics in medicine, explained in a press release that, because the system learns from the patient rather than requiring the patient to learn the system, the technology gives the patient much more control and produces results much more quickly than traditional AAC methods.
“Our [device] uses analogue signals in continuous form, which should give us a greater speed advantage because more information can be collected in a shorter space of time,” added Kerr, who reported that initial tests showed the system was 97.5 percent effective in learning “breath language.”
Because the device does not require muscle movements, Kerr believes it could give the power of speech to patients unable to perform the physical tasks necessary to make traditional SGD systems work, and it could even be a feasible option for patients on ventilators in intensive care.
An article published by the British Medical Journal emphasized that early diagnosis of locked-in syndrome, where a patient is completely physically paralyzed yet retains consciousness, was absolutely essential for patient recovery but often depends on care staff or family members to notice patient awareness.
Kerr noted that the Loughborough SGD could facilitate diagnosis of locked-in syndrome much sooner and give patients a better likelihood of regaining full or partial function.