Brain computer interface

Question

Do you believe brain computer interfaces (e.g. Emotiv Epoc, OCZ NIA) are good enough to be used in designing a useful user interface and if so could you please name some examples? Tnx.

Answer 1

Maybe in the far future if they get more powerful and less intrusive. But the tech just isn't there yet.

One of my friends final project for a hardware class was a brain controller for Mario. It was basically unworkable at first. After a few minutes, you could train yourself -- but you could never be as accurate with the BCI than with your hands. I think essentially the problem is that the brain mapping is just as arbitrary as the controller mapping, but much more complex.

Answer 2

I wrote my thesis about Brain Computer Interfaces for my master's degree in Human Computer Interaction.

My answer might be a bit long, but I feel it is necessary to give this relatively new topic (BCI’s) the necessary attention it deserves in the field of interaction design.

And this is my answer to your question:

Emotiv Epoc, OCZ NIA are considered to be non-invasive (not implanted) consumer based EEG devices which also filter out EMG signals (artifacts produced by your muscles and not by synapses inside your brain.)

Non-invasive devices can measure brain waves using simple electrodes attached to your head. The originating source of the signal (the part of the brain producing it) is hard to pinpoint. The signal gets distorted by background noise (interference from the outside) and also bounces multiple times against your skull. This means the signal from electrode E1 also contains brain waves already measured by electrode E2 and vice versa together with loads of other background noise.

This leaves us with a brain pattern which you can measure and even manipulate with some practice. But there is no way of saying that the measured signal is coming from that specific area of the brain. This means you cannot read someone’s "thoughts”. Brain patterns can be reproduced with some practice but there is no way of saying which pattern is linked to sadness or when you were thinking about sports with non-invasive consumer EEG-devices.

To make it even harder: a brain pattern is unique to the person producing it. So your device becomes very personal and you need to train it to recognize your patterns.

What can we do with it?

A user can train the device to recognize his/her patterns by producing a very specific brain pattern and store it into a personal profile.

The user can now manually create a command and attach it to the brain pattern so the system executes an action of some kind when he/she produces that same pattern again.

The Emotiv creators were smart enough to filter out and use the EMG signal from the initial unfiltered EEG signal. Usually these are considered as artifacts and are discarded. EMG signals CAN be classified. Therefore it is possible to use the Emotiv EPOC without training to detect smiles (pleased) or detect clenched jaws (signs of frustration), etc...

But should we use it as a way of interacting?

A user interface is considered to be useful when it feels natural to us and is easy to use.

Now think about this: is there anything physical in the real world you manipulate with your mind except your own body? (thanks Alex for the feedback, missed this)

Probably not, this is called [Psychokinesis]

I am sure many of you would like to (including myself), but this is not humanly possible as far as I know of. :)

Now ask yourself the question: would manipulating an interface with only thoughts feel natural to us as humans?

I don’t think so… But then again manipulating a GUI with mouse and keyboard isn't either. This is human nature, we adapt.

There is a different side to this story:

we are talking about explicit user interaction, where it is the user his intention to give out commands. It would be more useful to work with implicit interactions, where you analyze the needs of the user through his subconscious. Imagine detecting emotions like engagement, frustration proactively, even before the user experiences it himself.

Think about what it would do to the field of usability testing! No more guessing about what the user experiences when interacting with our interfaces.

This could also be potentially very dangerous as well: think about game companies designing games that dynamically adjust their difficulty level to keep engaged enough without losing our interest due to frustration.

We are simply not there yet:

There are other devices and techniques available today that can pinpoint the source of our brain activity more precisely. Unfortunately they are mostly used in medical institutions, research facilities and are not available for consumer use.

The consumer BCI devices available to us simply are not there yet. It’s a start and we should start experimenting with them immediately. However I believe using a BCI in real life for useful purposes is still part of the future. There are going to play a vital role in our lives but on a subconscious way of interacting.

After reading my answer about BCI’s, I encourage you to stop and think about how the field of interaction design will change when using a brain computer interface for active or passive interaction.

Answer 3

Did you see any demo with a non-trained user? On Discovery channel there was a series called "Prototype this" where in one episode they tried using a brain controller. I remember when the guy was trying it the first time.

The device works in a really simple way. They told him to think about a particular thing and the system took it as a sample of a command. So it didn't matter whether he was thinking about moving a box on the screen or buying candies. It had to be a brain activity that would be distinct enough to recognize.

Want to make Your own brain controller? Get a Infra-red cap that they use to monitor brain activity (sorry, I don't remember what it's called at the moment) and get a flat image out of it. Connect it to an artificial neural network and train it to recognize the images created while a person "generates a particular brain activity". And that's it.

Ok, so my point is this is useless as a controller and will remain useless for at least a decade. Using it to move a cursor on the screen would be as comfortable as moving it with cursor keys arranged in a row.

thank you.

Answer 4

No. Instruments like that are way overkill and no substitute to basic tenents of user interface design. IMO, testing your UI on a handful of users would be far more fruitful, and cheaper.

Answer 5

People want to use their hands. Not using their hands actually makes the user mad. Its because its in human nature to use their hands to do tasks. So, no. This makes life less enjoyable for the user if there is nothing to touch or interact with physically.

Answer 6

There's an edition of the UK TV programme 'The Gadget Show' where the team (at least appear) to use a BCI to pick up a car with a crane and move it around a warehouse.

Setting a world record for the biggest object to be used with a BCI

Here's a link to the technology:

http://neurogadget.com/2011/04/12/neurosky-mindwave-sets-guinness-world-record-for-%E2%80%9Clargest-object-moved-using-a-brain-computer-interface%E2%80%9D/1820

And here's a description of how well things went:

http://www.ag-crane.com/the-gadget-show.htm

Answer 7

One of the problem is that it's really hard to control your thoughts (it's human nature). There are some mental practices like "stopping your inner dialogue" and others but it's really very hard to learn to control you brain. And if you can't control your own mind, it's extremely difficult to control anything else with your thoughts.