Last week I was able to attend h2.o, a conference organized by the MIT Media Lab that was focused on Human 2.0. The theme of the conference was new minds, new bodies and new identities and to help support the goals of the them the Lab has recently created a Center for Human Augmentation.
I was interested in attending after hearing Frank Moss, the director of the Lab, speak at an MIT Communications Forum event earlier this year. While I often attend events because of a personal curiosity or professional aims, this one was different. My son, who is 10, has a number of neurological and psychological issues. His condition has profound effects on our family.
When I arrived at the conference last week I was running late. I missed the keynote by Oliver Sacks and some of the first session. That was a bummer but not the biggest deal in the world. The first session wasn’t what I wanted to hear (although it was interesting none the less).
I was interested in the morning’s second session – which was on the theme of “new minds.” There were three stand alone presentations – by Ed Boyden (assistant professor of media arts and sciences, MIT Media Lab), Douglas Smith, MD (professor, department of neurosurgery and director of the center for brain injury and repair, University of Pennsylvania) and John Donoghue (Henry Merritt Wilson professor of neuroscience, Brown University).
Ed Boyden – Engineering the Brain: Toward Systematic Cures for Neural Disorders
Boyden is working on a project focused on re-engineering the brain’s circuits. The lab is new, as are the projects. The hope is to develop new tools to treat the brain directly.
The goals of the projects are:-
To treat neurological and psychiatric issues
To augment cognition
To better understand the human experience
Doing these things requires new, systematic tools. The 20th century was the era of pharmacology with specific drugs available to solve single specific problems. Boyden hopes to create solutions that can be used to address multiple problems.
The challenging thing, of course, is that the brain is really complicated – and as you zoom in the complexity grows and grows. So how does one apply engineering concepts to these complex systems? By looking at behavior.
Boyden went on to discuss three potential approaches they are working on:
The first are devices for non-invasive brain stimulation. These are safe, can turn on or off specific regions of the brain and are being tested to treat conditions like depression. A wearable version is currently being developed in the Lab. In addition, there is also work being done on more focused – but still non-invasive – stimulation technologies.
The second is engineering software for automated, customized adaptive therapy. Boyden was looking at hypnotherapy and noticed that the scripts looked a lot like computer programs. They are using this similarity to develop customized hypnotherapy scripts. He demonstrated this with the conference moderator, John Hockenberry. The system asked a series of questions which modified the phrases and flow of the script. Boyden pointed out that while this customized hypnotherapy can be used to relax, it is also designed to help people develop or strengthen cognitive skill.
The final approach he presented was the ultra precise engineering of neural computation by optical neural control. This was the most gee whiz of the three. Basically, the idea is to use light to trigger and control neurons with the goal being the creation of optical neural control prosthetics. They understand how this works and will next focus on prosthetic design. Boyden can see applications for this approach in dealing with conditions like blindness, deafness and Parkinson’s. Lest this all be thought of as pie in the sky, he pointed out that this is being tested now.
It was exciting to see and hear about the different approaches being considered, studied and developed at the Lab.
Douglas Smith, The Brain is the Client: Designing a Back Door into the Nervous System
Smith is at Penn and is focused on traumatic brain and spinal cord injury. He described progress in developing a brain-machine interface as being at a crossroads as people try to figure out how to move electrical signals to and from the brain. For example, where and how does one connect to the brain?
As far as Smith is concerned a hard/sharp interface is not a good idea. The nervous system, he explained is promiscuous and so a wet and juicy interface (which is what it is used to) is more appropriate. But, he wondered, do we need to connect directly to the brain at all? He does not believe that we should and suggested establishing a brain/machine interface as far from the brain as possible to take advantage of the processing power of the central nervous system.
This power needs to have true, two-way communication if it hopes to do allow for complex tasks. Simple on-off functions are not good enough. Getting signals in and out – and allowing for the performance of complex tasks means that at some point there does need to be a hard interface. Smith and his team are addressing this with nerve development – the creation of neurofilament. Doing this means that there must be growth of neurons and axons.
Smith is developing a means to stretch axons to allow them to span multiple neurons. Nerve fibers obviously grow, but no one is quite sure where the growth occurs. This stretching technique appears to be working and seems to be mimicking natural processes.
The axons that have been grown in culture can then be removed and used to bypass damaged neurons and connect healthy isolated ones. Axons can grow fast and they are harnessing the capacity of axons to grow to address traumatic injury. Can they be used to repair spinal cord injuries as great as three centimeters? Smith showed images of this being accomplished.
Another application he discussed was the creation of a nervous tissue construct. In this case they created tube of nutrient/culture for the axon to grow into and around and were then able to transplant the resultant nerves into a animal. After four months, the new nerve was integrated into the animals nerve network.
It is not a huge leap to apply this to approach to lost limbs by connecting a multi-electrical array on a device to a host nerve using the grown axon package. This allows a wet-to-wet connection to a hard/tensioned device.
They have recovered neurons from patients and organ donors; and these neurons can be preserved and used to seed and stretch axons which is leading to real and practical clinical applications. They’ve figured out how to integrate with the nervous systems, how to move an electrical signal back and forth across the axon for two way communication. The only missing piece is the device able to receive and respond to the signals and that, Smith believes, will come.
John Donoghue, New Successes in Direct Brain/Neural Interface Design
Donoghue is known for developing brain/machine interfaces and devices. He is the executive director of Brown University’s Brain Science Program and is the founder and CTO of Cyberkinetics.
He discussed the use of neural interface devices that could be coupled directly to the nervous system to diagnose problems, treat conditions and repair function. Such devices already exist, he pointed out, using the pacemaker as an example.
Today, the ability exists to get signals into the brain (through electrical stimulation) and back out (with sensors). Neurotech is here – devices like cochlear implants and early work in retinal implants as well. Another – deep brain stimulator for movement disorders
At its simplest level, Donoghue described the human nervous system as being the brain sending a signal to a muscle resulting in an action. There are a number of conditions that can break the connection between the brain and the muscles. To deal with this, systems are being designed and developed to reconnect stranded brains to the outside world. They all consist of a sensor (that receives the brain’s signal) and a decoder that receives the signal and converts it into a signal that a device can react to.
CyberKinetics is working the BrainGate interface system. It consists of a hard interface (a 100 microelectrode array) that connects directly to the cerebral cortex. The signal processing is done externally and then sent to a system or device. There are currently four people using BrainGate – two w/ spinal cord damage, one with a deep stroke and one w/ ALS.
To make the system work, Donoghue and his team first needed to determine whether there were still signals occurring in the motor cortex (after the insult or injury causing the patient’s condition). They next had to modulate the signal by asking the patient to “preform” the task to see the signal of the intention to make the movement. Their ability to do this demonstrated that the signal was there and that it could be harnessed to control devices. Donoghue went on to show a number of videos of patients controlling movements and devices with the BrainGate interface.
The next steps for this technology are the development of a wearable system (for which a prototype already exists) and a system for connecting BrainGate to muscle for movement control. Over the next five years, Donoghue expects a whole array of sensors and stimulators that will be able to address a range of conditions.
While none of the three presentations was a silver bullet, it was heartening – both as a parent and as someone who believes in the positive possibilities of technology – to see smart people applying novel thinking to solving pressing problems.
[tags]MIT, Media Lab, H2.0, h2.o, Frank Moss, Ed Boyden, Douglas Smith, John Donoghue, Cyberkinetics, BrainGate, neural disorder, nerve growth, brain/machine interface, brain, spine[/tags]