Brain computer interfaces (BCIs) are a link between the brain & an electronic device and are utilized in research, repairing functions for patients suffering disabilities such as blindness and perhaps in the future for augmentation.

Inspired by a chapter in a book called More than human by Ramez Naam I recently gave a talk at BuzzConf on an overview of BCIs and was amazed by just what is possible - this stuff is no longer (and hasnt been for some time) the subject of cyberpunk novels.

Before we start my background is software development & I’m certainly no expert in BCIs.

I also want to mention that BCI research often utilizes animal experimentation & this at times makes for some uncomfortable reading at best. The subject of this post is not to debate the ethical implications of this.

Why develop brain computer interfaces?

So why on earth would we want to plug our brains into a machine?

Well there are many reasons we might want to do this such as:

  • BCIs have the potential to drastically improve the quality of life for those suffering disabilities and illness such as blindness & nerve system damage
  • To understand our brain & nerve system better
  • Create new interfaces to devices e.g. think about controlling vehicles, devices & computer games by your mind alone
  • Enhancement e.g. improved vision, overlayed displays etc

BCIs of course have implications for many areas such as the military and even marketing - I don’t think it’s too outrageous to think of adverts that change depending on how you are responding to them which is a bit er scary..

A bit of background

We are not too sure how the brain works but we do know that activity in the brain generates an electrical field that can be measured.

As a species we have known about this for some time with written evidence that Roman court physicians treated headaches using Torpedo fish better known as electric sting rays - how this discovery was made is anyone’s guess!

The brain is made up of several areas & we know by studying this activity when paired with stimulus areas seem to be associated with different functions.

For example:

  • Frontal - Reasoning, planning, speech, movement, emotions
  • Parietal - Movement, orientation, recognition, perception
  • Occipital - Visual processing
  • Temporal - Perception and recognition of sounds, memory & speech

We have a guy called Hans Berger to thank for coming up with a device to measure electrical activity in the brain.

Hans developed the first EEG (electroencephalogram) device which was a little more primitive than what’s available now and involved silver wires being inserted under patient’s scalps. An EEG measures electical activity at different frequencies

Hans became fascinated in the brain after having an accident in a military training exercise. Many miles away Han’s sister experienced a feeling that her brother was in trouble and insisted her father send a telegram to check he was OK & Hans became very interested in the idea of a psychic link.

Back to brain computer interfaces (BCIs)

Anyway back to BCIs - there are two main types of Interface - invasive & non-invasive & some things that sit somewhere in between.

Non-invasive options include:

  • EEG
  • fMRI
  • Magneto-encephelography (MEG)
  • Near infra-red spectroscopy (NIRS)

Invasive options include various types of implants.

Some of the above options might be familiar as are often used in to treat & identify various medical conditions such as tumours, strokes and epilepsy.

Non-invasive options have the following advantages:

  • Easy to use – it’s a lot easier to wear some headgear than have something implanted directly in your brain!
  • No infection or rejection risk with non-invasive options
  • No ethical issues with installation

On the other hand non-invasive approaches have some disadvantages:

  • Poorer resolution of signal – you are just not as close to the source of the signal & its more prone to noise
  • Impossible to detect some changes deep in the brain
  • Limited portability e.g. you are not going to be lugging an MRI scanner around

Invasive approaches have pretty much the reverse of the above in terms of advantages & disadvantages.

Key Studies

There are a large number of studies in the field of BCI & I have picked out a few - I’m also limited by what I could find without subscribing to various academic journal sites so suspect there are many more.

José Delgado (1965) implanted an electrical device in bulls brain that he claimed would cause it to lose its aggressiveness when activated. Jose would annoy the bull, wait until it charged at him before activating the electrode to make the bull passive.

You can watch a video of this at https://www.youtube.com/watch?v=23pXqY3X6c8.

As early as 1973 Jacequeus Vidal demonstrated that subjects could move a “cursor” around a checker 8x8 checkerboard using just recordings from the visual cortex.

Kennedy and Yang Dan (1999) decoded images from a cat’s visual cortex. There is a video (with an unhappy looking cat) that you can watch on this at https://www.youtube.com/watch?v=vGNvP75UYGE.

Donoghue worked with Rhesus monkeys and using brain signals to control a robotic arm.

A reseatcher called Dobelle (1978) had success with an implant to provide limited vision to a patient called Jerry. In 2002 Dobelle worked with a patient called Jen’s & had great success with a device that allowed Jens to drive a vehicle slowly around a carpark. Dobelle was unusual in that his research was privately funded.

Matt Nagle was a tetraplegic after being stabbed and with the aid of a BCI controlled an artificial hand. Sadly FDA regulations demanded device removed after 1 year. It must have been devastating to lose the advantages given by the device I guess no one knows about the long term impact an implant will have.

Nenadic & Do (2015) had success with a paralysed participant who walked 3.66m using just an EEG device.

Gallant et al (2011) had success with reconstructing visual imagery by presenting a series of inputs & measuring the brains response to these input. By having a large set of information on how the brain responded to various inputs this could then be used to reconstruct the input based on the brains responses. There is an awesome video of this you can watch at https://www.youtube.com/watch?v=6FsH7RK1S2E

Nirenberg (2015) had a similar approach to decoding neural signals (these things don’t come with a manual!) by taking the approach that you might not know how code works - but you can work backwards.

By recording electrical pulses to known inputs a computer to work out these signals translate by using brute force. Check out Nirenberg’s TED talk.

Consumer Devices

Believe it or not you can experiment with BCIs yourself - in a non-invasive manner of course!.

At the time of writing there are two main consumer options: the NeuroSky mindwave & the Emotiv.

The NeuroSky Mindwave retails as $79.99 USD and has very limited capabilities. I have one of these and although a bargain for the price after the initial novelty wears off there is little you can do with it.

A better (& much more expensive option) is the Emotiv. The Emotiv has a lot more functionality & retails at $400 USD+ depending on which option you go for.

Why not have a play with one of these & who knows you might discover something new?

Further reading: