I first learned about the Bionics Institute when reading IEEE President Karen Bartleson’s December column, in which she noted some of the incredible, real-life applications she discovered during her visit there this year.
The institute, in East Melbourne, Australia, is a leader in hearing research, particularly cochlear implants, surgically implanted electronic devices that provide a sense of sound to people who are deaf or severely hard of hearing. The institute is using that same technology to improve vision and monitor brain activity.
I interviewed the institute’s CTO, IEEE Fellow Hugh McDermott, a pioneer in cochlear implants, about several of its latest projects.
A BETTER FIT
Each cochlear implant must be fitted to the individual and her comfort level of loudness, which can be difficult to gauge. Fitting patients who are as young as 6 months is especially complicated because babies cannot describe what they are hearing. The institute has developed a new type of brain imaging technique called functional near-infrared spectroscopy (fNIRS), a noninvasive way of seeing how the brain responds to sounds.
With fNIRS, the patient wears a skullcap that’s similar to an electroencephalogram (EEG) cap, embedded with optodes. Optodes have light sources that operate in the infrared region and detectors that pick up light that is scattered back out of the brain. fNIRS measures the change in blood oxygen levels in each region of the brain, showing how active that part is. fNIRS imaging allows researchers to see how the brain is responding to the hearing device and tailor it to individual needs.
“If we could get an objective measurement of what’s actually happening in the brain when the cochlear implant recipient is hearing a sound, then we would know how to adjust the device and get the best possible fitting,” McDermott says.
The institute is conducting proof-of-concept clinical trials. Commercializing the technique is still a few years away, he says.
The Bionics Institute began developing a device in 2007 that used an electrode array to stimulate the retina of patients with retinitis pigmentosa, the most common cause of inherited blindness for which there is no effective treatment. For those who have the rare eye disorder, the light-sensitive cells of the retina—the photoreceptors—degenerate, but the retinal neurons that transmit information to the brain remain.
The device electrically stimulated the receptor cells in the retina, and bypassed the receptors that were lost or damaged. By stimulating the cells, visual sensations in the brain can be created, McDermott says.
The electrodes were placed near the retinal cells and connected to an electric stimulator, similar to the stimulator used in cochlear implants. Each electrode produces a spot or flash of light. During a clinical trial using the prototype device completed in 2014, patients reported “seeing” a type of connect-the-dots picture.
“Patients’ normal vision isn’t restored by any means, but they do get information about things, like where the door is located or the placement of a table or chair,” McDermott says. “It helps them navigate.”
The institute has now developed a more advanced device with additional electrodes and a stimulator placed under the skin, similar to a cochlear implant. The system includes a small camera worn on eyeglasses and a video processor that can be carried in a pocket. The processor extracts information from the camera’s field of view and picks up obstacles in front of the person. Information about an obstacle is encoded into a pattern of dots and transmitted to the implant. The stimulation of the retina produces a dot-pattern image that helps the patient identify shapes, perceive movement, and navigate their way around without assistance, McDermott says.
Three patients will receive the implant next year.
ADAPTING TO ABNORMAL MOVEMENTS
Deep brain stimulation treatments have been used for several years to manage Parkinson’s disease, a progressive disorder of the nervous system that affects movement. Electrodes are implanted into parts of the brain related to movement, and the electric stimulation continually pulses the area to remove or reduce symptoms such as tremors and rigidity. The treatments become less effective over time and although the symptoms constantly change, the stimulation can’t be adjusted easily or quickly enough to maitain optimal therapy, McDermott says.
The institute has developed a way to modify the stimulation in real time to adapt to the changes.
“The Holy Grail of our project was to find a signal that varies with symptoms, and we think we’ve found it,” McDermott says. “We can now measure the signal through the very same electrodes that are used to deliver the stimulation. This is a big step forward.”
That same signal also could help surgeons implant the electrodes with more accuracy. The implant procedure is difficult because the surgeon is working with minimal information, according to McDermott.
The electrodes have to be placed deep in the brain, and “it’s hard to visualize where with current imaging techniques,” he says. “What the surgeons need is to have a signal like a lighthouse that tells them where the target is, or to move the electrode closer. The same signal we think can control the stimulation could be used to help get the electrode to the right place.”
Studies are being conducted with patients who have existing devices, he says, adding that clinical trials are still a few years off.