Kennedy said his decision to experiment on himself followed frustrations that impeded his own research. His company, Neural Signals, was working on a software speech decoder that would translate brain signals into spoken words, but the project stalled after he failed to receive FDA approval for future experiments. Kennedy also was frustrated by the data he obtained from his severely disabled patient pool. Because they often were unable to communicate, it was difficult to confirm what they were thinking when a neuron fired, a critical piece of information he needed if his experiments were to be successful.
Kennedy decided that he needed to expand his experiments to include volunteer patients who could speak. This would allow Kennedy to confirm that his decoder was accurately interpreting what they were thinking. After searching a year for a volunteer with ALS who could still speak, Kennedy gave up. “I couldn’t get one. So after much thinking and pondering I decided to do it on myself,” he says. “I tried to talk myself out of it for years.”
Kennedy took matters into his own hands and paid for surgery to install electrodes into his own brain. Since US doctors could not perform such a surgery, Kennedy traveled to Belize, where doctors agreed to perform the surgery for $25,000. The 12-hour surgery took place in June 2014, but it did not go as smoothly as planned with Kennedy initially losing his ability to speak. He followed his first implant surgery with a second surgery several months later to implant the electronics to collect the signals from his brain.
After surgery was complete, Kennedy began his self-experiment in earnest recording brain signals both while he spoke and while he imagined speaking. He focused on basic phonemes and simple words and discovered that the different combinations of neurons that fired when he spoke also fired when he thought about speaking. This relationship is critical to developing an accurate speech encoder and its discovery was a major breakthrough for Kennedy.
Unfortunately, Kennedy’s experiment had to be cut short, lasting weeks instead of years, because the incision in his skull never completely closed. He had to pay to have the implant removed at a local Georgia hospital. The neurologist reported his early findings at the annual Neuroscience 2015 meeting held last month in Chicago, where his experiment was received with both awe and concern by his colleagues.
