These days, it seems you’re nobody if you’re not working on a way to merge machines with the human brain. Earlier this year, both Facebook and perpetual moonshot-enthusiast Elon Musk announced plans for brain-computer interfaces that could allow us to read the thoughts of others and improve our capacity for learning. Today, the US Defense Advanced Research Projects Agency announced plans to spend $65 million developing advanced neural implants that connect our brains to computers in order to treat sensory deficits like blindness.
The Neural Engineering System Design programme gives us a peek into what sort of achievements might actually be plausible using neural interfaces. The programme will fund six different research teams, including two that seek to restore vision using light-emitting diodes, one that plans to decode speech using “neurograin” sensors, and another that uses holographic microscopes to detect neural activity that could eventually replace lost vision, or act as an interface to control an artificial limb.
At the end of the four-year programme, the goal is to have working prototypes capable of transmitting data between the brain and computers, but it will likely be a good deal longer before such devices are ready or commercial or clinical application.
“It will be a long time before medical science allows us to grow new eyes or repair a broken spinal cord, but by linking brains to computers it will be possible to leverage digital devices to restore the functionality of damaged body parts,” said Matthew Angle, whose company, Paradromics, Inc., received a DARPA grant. Angle’s company is researching how to use large arrays of brain-penetrating microwire electrodes to record and stimulate neurons, with the goal of building an implantable device to support speech restoration. His company aims to be in clinical trials by 2021.
“Initially we are focused on what we call ‘connectivity disorders,’ meaning illnesses and injuries that destroy or severely impair a person’s sensory or motor connections to the outside world,” he told Gizmodo.” Looking forward, I imagine that neural prosthetics could also be used to treat certain neurological diseases.”
It’s not DARPA’s first foray into brain-computer interfaces. The agency has already invested heavily in brain-computer interface technologies to do things like treat mental illness and restore memories to soldiers injured in war. (Those projects are still ongoing, but so far on track to meet research goals.) But the goals—and the technology—here are a little different. Rather than seeking to impact one small region of the brain in order to affect a particular outcome, like treating PTSD, the agency hopes to develop a technology to communicate with more than a million of the brain’s 86 billion neurons at once, translating the brain’s electrochemical signals into ones and zeros that can be interpreted by a machine.
Such a feat would have countless therapeutic applications, but also significantly expand our understand of vision, hearing, and speech and eventually, yes, maybe even allow us to communicate telepathically.
At UC Berkeley, for example, a research team led by Ehud Isacoff plans to develop a holographic microscope that uses fields of light to detect and modulate the activity of up to a million neurons in the cerebral cortex. The team hopes it can create models that predict how the brain will respond to visual and tactile stimuli, then translate those models into patterns that might convey vision or movement to someone who had lost one of those senses with a brain implant.
“The technical goal is to create a brain modem that can ‘read’ the activity of a million identified neurons and ‘write’ back to large numbers of them patterns of activity that simulate natural ones,” Isacoff told Gizmodo via email.
Using optical imaging, he said, may be more effective than techniques like deep brain stimulation that rely on implanted electrodes to stimulate areas of the brain around them, allowing to scientists to target extremely precise regions. Within four years, he said, they hope to have a device that works in animals.
“We hope that our device makes it possible to unlock the neural code of sensory perception,” he said. “Success would enable us to generate the proper patterns to reflect what is happening in the world to enable a blind person to see or someone with a prosthetic arm to control it better because of restored sensory feedback.”
At Columbia University, a team led by Ken Shepard plans to create a prosthetic restoring sight to those that are blind, by layering a single, flexible circuit over the brain that could communicate wirelessly with a transceiver worn on the head.
“The goal of this research is to push the limits on what is possible with brain-machine interfaces—providing a means to interact with brain circuits on a scale that has not previously been achievable,” Shepard told Gizmodo.
The challenges, though, are many. How to make such a device survive inside the body? How to process the data? How to map signals from the brain and understand how they impact the brain’s complex wiring?
DARPA’s goal is that all of the teams will eventually create technology with practical, commercial applications.
Angle, of Paradromics, cautioned that this does not mean we will all be reading each other’s minds anytime soon. Even twenty years out, he suspects we will still be puzzling over how to use this kind of technology to help people with physical and psychological illnesses. In the nearer term, though, the focus will likely be on disorders rooted in the brain’s inability to communicate with the body, like blindness and paralysis.
“There are already enough medical applications to keep many companies busy for many decades,” he said. “We see a concrete and credible technical pathway for the blind to see and for people who cannot move to walk again. This has been a human aspiration for as long as written history, and I think the tipping point will come in the next decade.”