Brain Training

Thought Beanie apps enable brain training. Here's how.

Brain Training

One of the key elements of Thought Beanie apps are exercises that train the user's brain to adopt desired states or habits. This could improve concentration and focus in sport or help get to sleep. Thought Beanie applies neurofeedback training techniques.

The process of Neurofeedback Training (NFT) provides individuals with explicit real-time information about a) their current neural state and b) desired state. The difference between these two states represents the amount of error in their behaviour (at a neural level) at that precise moment in time.

How does NFT work?

There are two popular ways of providing this information—through auditory and visual feedback. The NFT theory goes that by seeing or hearing this error feedback, over time, people should be able to learn the mapping between their brain rhythms and the error signal and subsequently adjust their activity so error is reduced (this is the fundamental principle behind the majority of learning processes).

On a longer timescale, most likely on a scale of several months, you should be able to control these rhythms without the external stimulus. Crucially, however, you will never get to that stage if you don’t learn the relationships first between input and output – why it’s training!

A Smartphone app for NFT?

When you open the app, we ask for the desired level of difficulty of the game, essentially the degree of the difference between the baseline (explained next) and the desired state.

Brain Training Base State

We next record people's baseline states of alpha and theta (the brain rhythms we are interested in modulating) at eyes open and closed. These measures (the mean power of the signal for each frequency band) provide a starting point for where you are currently at. If this was a fitness training app, it would sort of be the equivalent of taking a record of your height and weight and current activity levels.

The actual calculation of difficulty (set above) is based on the number of standard deviations (SD) away from this mean baseline activity. So a SD of 1 means you are looking for ~68% improvement and a SD of 2 is ~95%. We want to go with small numbers in the beginning and build up to bigger improvements. We could do this in an adaptive manner (and will no doubt want to at a later stage) but for now we manually set it at each session.

stencil_golfer-drive_0dadce13305aa9d304d0d70b26839955_800.jpg 35.58 KB

A Brain Training Trial

Once this target value/difficulty level is set, we start a trial. A trial is just a period of time when the screen shows a transformed measure of real time EEG activity – where mean alpha power is represented as a blue circle. The black border- empty circle on the outside of this blue circle indicates the desired state (the larger the circle relative to the blue line, the more difficult the task).

The aim of this game is to try to fill up as much of the black border- empty circle with the blue inner circle. The only way to do this is to increase your alpha power. We have a running score at the bottom which is the root mean error difference between the two circles- a quantification of the degree of error.

We have only set up one version now for illustrative purposes. A complete training protocol would ask you to try to decrease your alpha, too (and increase and decrease theta too).

Brain Training Games

The exercises we’re working with at the moment are pretty basic, but increased gamification and complexity will follow, even integrating into existing games. The current state signal could be used to trigger the pull on the slingshot on angry birds for example – where success is visualised as number of eggs retrieved but still is a representation of the difference between the current and desired state.

stencil_vr-headset_f6b333952f082889611f44c1a01f4d49_800.jpg 29.52 KB

Before we start going into such degrees of complexity though, the circles and dots seem like a safe place to start…