Tiny Neural Prosthesis Could Help to Restore Vision and Hearing in Humans
FlatScope would be implanted between the skull and cortex in order to communicate with functioning areas of the brain that can process sight and sound.
In January 2016, DARPA announced the $65 million Neural Engineering System Design (NESD) program, with a goal of developing an implantable system to provide precision communication between the brain and the digital world. It is now known what that system will look like, as Rice University engineers recently unveiled a prototype for what they are calling FlatScope — a small, flat microscope that could be implanted between an individual’s skull and cortex.
Ashok Veeraraghavan, an associate professor of electrical and computer engineering at Rice, explained, “The goal is to place FlatScope directly on the surface of the brain.”
His colleague, Rice engineer Jacob Robinson, added, “Certainly one application is to better understand how the brain processes information. For the DARPA program, we are specifically focused on building a better neural prosthetic.”
Similar to how an artificial limb handles functions once achieved by a missing leg or arm, FlatScope might compensate for lost vision or hearing by delivering digital information directly to parts of the brain that can process it.
A DARPA announcement further describes the goals for the implantable device. It mentions that the analysis of brain processing should be done with "individual neuron-level precision" and "at a scale sufficient to represent detailed imagery and sound." Once those biological processes are sufficiently understood, then the selected research teams are charged with developing "strategies for interpreting neuronal activity quickly and with minimal power and computational resources."
A cousin to FlatScope is FlatCam, which was funded by the National Science Foundation and the Office of Naval Research. FlatCam has already been in development for a few years. This lens-less camera is ultra-small and is thinner than a coin.
“We have demonstrated utility of FlatCam for several low-cost, low-power consumer imaging applications and are continuing to develop the fundamental technology and the specific application scenarios, such as face detection,” Veeraraghavan said. “We are also currently in the process of finalizing joint projects with industry evaluating the applicability of FlatCam in specific application scenarios.”
The researchers are incorporating FlatCam’s advanced imaging abilities into FlatScope. Over the next four years, and with $4 million of the NESD funding, they will develop an optical hardware and software interface for the device. This work will be in collaboration with the Yale University-affiliated John B. Pierce Laboratory, led by neuroscientist Vincent Pieribone.
During the initial phases of research, when the researchers plan to study brain processing in mice, the optical interface would detect signals from neurons modified to generate light as they are active. This feat would be achieved by programming neurons with proteins that release a photon when triggered, mimicking the bioluminescence of creatures like glowing jellyfish and fireflies.
Veeraraghavan said that FlatScope would be able to capture 3D images, permitting the scientists to view not only neuronal activity at the brain’s surface, but also an estimated 500 microns deep into the brain’s cortex tissue. This deeper region of the cortex is where the scientists believe most visual and sound processing occurs.
“Currently, power and data (in FlatScope) are transmitted with wires just like the camera in your phone,” Robinson said. “Our goal in collaboration with our team members from Ken Shepard’s lab at Columbia is to create a completely wireless method to deliver power and data.”
Ken Shepard heads up Columbia University's Bioelectronics Systems Lab, which specializes in solving such technical challenges. The lab, for example, holds patents for everything from a DNA-based thermometer to a system that provides energy-efficient, low-voltage power to semiconductor circuits.
The FlatScope team is aware of concerns about the potentially harmful radiation released by devices such as phones and headsets that are often placed close to a user’s head.
Robinson, though, said: “As part of the DARPA project, we will work with the Columbia group to carefully design the data and power interface within the guidelines for what is safe for the human brain.”
His colleague Caleb Kemere, who is an assistant professor of electrical and computer engineering at Rice, said demonstrating the safety of the technology is a major challenge.
“While a number of optogenetic tools have been demonstrated in rodent and primate models over the last decade, they are only beginning to be used in humans,” he said. “While so far there have not been any reports of side effects, this is probably the primary time-limiting challenge.”
Yet another challenge will be to eliminate the need for the device to be hooked up to a larger, more powerful computer.
“In the future, we expect that each FlatScope would have its own embedded system to process the images and apply a neural stimulus,” Robinson said. “The first clinical applications will likely be neural prosthetics that can restore lost senses like vision and hearing by measuring the activity in a sensory-related brain region and applying a stimulus directly to that cortical area.”
Few would question goals of restoring sight to the blind or hearing to the deaf. Similar technologies could also possibly be used in the future to develop new treatments for Parkinson’s, Alzheimer’s, and any condition that robs an individual of some form of brain function.
Robinson explained that current probes that monitor and deliver signals to neurons for those with conditions like Parkinson’s and epilepsy are extremely limited. Even state-of-the-art systems have only 16 electrodes, which, he said, “limit how well we can capture and represent information from the brain.”
In contrast, the new system under development — inspired by semiconductor manufacturing — would have a thin interface that could monitor and stimulate possibly millions of neurons in the cortex. That amount represents just a small fraction of neurons with the brain, but it would nevertheless mark a biomedical breakthrough.
In the right hands, such work holds promise for improving the lives of countless individuals worldwide, including those with paralysis. As Kemere said, “DARPA has previously funded a number of programs focused on reading out neural activity to restore motor function in patients with partial or total paralysis. The FlatScope system also has the potential to revolutionize these sorts of prosthetic interfaces.”
The idea of a manmade device controlling a human brain in part or in full — particularly in the wrong hands — brings up ethical questions, however.
“Security is a major point of emphasis for this and any other biomedical implant that communicates with the outside world,” Robinson said. “DARPA recognizes this fact and, as part of our development plan, we must demonstrate that our system protects against unauthorized access.”
WATCH: Rescue Robots Battle It Out at DARPA Challenge