Turning brainwaves into sound

Image from news.Discovery.com

My best friend in college gave me Tuesdays with Morrie while we were in our third year to keep me entertained as I was traveling about visiting graduate schools. I remember nearing the end of the book while sitting in the back of a 737, sobbing so uncontrollably that a flight attendant came to check on me.

This book launched me into the stark realization of the existence of ALS (amyotrophic lateral sclerosis), also known as Lou-Gehrig’s disease, named after a famous Yankees baseball player who was diagnosed in 1939. This disease affects the central nervous system and ultimately leads to complete paralysis of the body while the brain remains otherwise unaffected and lucid. This condition of complete physical paralysis combined with intact cognitive ability is known as locked-in syndrome and affects approximately 50,000 Americans as a result of injury or a disease such as ALS. Frank Guenther of Boston University and his team have developed a system that uses an EEG cap to read brainwaves through the scalp to move a cursor on a computer screen (see here). When the cursor moves into one of three circles, the computer will produce either an “uw”, “aa”, or “iy” sound. While this is still far from producing actual speech, it is the start to giving locked-in patients the ability to interact with society again. They can also use a similar system to move a robot.

Interestingly, back in 2009, the team also tested a more invasive device that involved implanting an electrode in the brain and using a computer system to interpret the patient’s thoughts. They actually performed this procedure on a 26-year old who had been paralyzed by a brain stem stroke. A few months after implanting the electrode, nerve cells grew into the electrode and produced detectable signals. Once the team developed software that could detect the elements of speech through all the “neural noise”, they were able to help this patient reproduce speech with an approximate 50 millisecond delay — a speed comparable to what we typically take when speaking naturally! See more on this study here.