On the picture above you see graphs of alpha wave levels (relaxation) and beta wave levels (focus) from all electrodes on the headset. You see left and right for alpha and then left and right for beta on each row. See my previous post on brain functions to understand the positions on the scalp from which these recordings were taken. Basically the order from top to bottom reflects the orientation from front of head to back of head. This experiment consisted of making a phone call and have a conversation while registering my brain waves.

On the plotting of data from the left prefrontal electrode you clearly see a decrease in alpha wave values corresponding with a block of sustained peak values in beta (blue circles). This is quite an extreme example in the recording of cognitive concentration to perform the task of conversing, formulating questions and answering to the other party’s questions. The alpha waves show a majority of values under 10%, whereas the peak values in beta show values over 30% (percentages are measured differently for beta than for alpha but in my plotting all percentages are scaled respectively to a range between 1 and 10). Alpha values for the right prefrontal area are inconclusive with values centering around an average. The beta values for the same area however (top right window), are very significant in a totally different way from peaking at maximum. In this window you see how a repetitive pattern of sustained identical values is established. This pattern also shows when the test subject is writing or reading e-mail. This type of block iteration can be monitored on several windows and can be considered significant for the occurrence of an event, which in this case means the act of getting information – while talking on the phone to my wife – about how my daughter is doing in daycare.

Another striking data profile is found on the occipital electrodes at the back of the head where the visual cortex resides (green circles). Again alpha is decimated, especially on the right channel, and simultaneously the beta window shows sustained peaks (and a lot of separate peaks on the left side). The orange circle shows you there’s a difference between left and right vision, since peaks start with a different timing. At present I don’t have enough test results to make conclusions as to why beta values for visual processing are elevated here.

FP1 covers Brodmann area 10 on the front left side of the frontal lobe. This is part of the prefrontal cortex, that highly developed part of the brain which sets us apart from other mammals since it is responsible for the execution of cognitive tasks. Complex behaviours and simultaneous mental activities need a kind of working memory that keeps track of running tasks in either pending states or executive states. For most cognitive functions information needs to be retrieved after completion of another task. The prefrontal cortex co-ordinates all of this mental traffic and shows clear elevation in beta levels when performing calculative tasks, logical puzzles or other intellectual questions. FP2 covers the right side of the prefrontal cortex.

F7 and F8 cover parts of the frontal cortex. The frontal cortex is involved in integrating sensory information with data retrieval from memory locations. This way new sensory information can be compared with earlier perceptions. Area 45 – which is one part of Broca’s area, a brain centre dedicated to language production – at position F8 is also believed to recover semantic information and to evaluate that information in the light of the current context.

C3 and C4 are located on top of the primary somatosensory cortex, in the parietal part of the cortex. This is a clearly defined strip on top of the brain responsible for processing touch and sensation as well as keeping track of the location of your body parts (proprioception).

The two electrodes at the back of the headset, O1 and O2 for the occipital area of the cortex, cover the secondary visual cortex which is processing information relating to visual association. Cells are tuned to simple properties such as orientation, spatial frequency, and color. Depending on the exact positioning of the headset – which is likely to be slightly different per individual – they might also track the primary visual cortex from which the information is forwarded to the secondary visual cortex. Primary cortex here corresponds to Brodmann area 17 and the secondary to area 18. The connectivity between primary and secondary cortices is important for visual memory.

Table to check which Brodmann area corresponds to which electrode position

musical interface for Staalhemel

Preparing the musical interface in Max/MSP to control the Staalhemel installation as a percussive apparatus. On the opening of the Sound Art expo at Festival van Vlaanderen Kortrijk the premiere will take place of this concert version. In the concert version De Boeck will operate the 80 steel tiles manually via a touchscreen while his brainwaves will operate other musical parameters.

Concert:  24 april, Budascoop, Kortrijk (B) – 20:15

Extended electrode 10-20 positions

A definite selection of the electrode positions has been made. IMEC hardware now allows for transmission of up to 8 channels simultaneously. Electrodes for the following positions will be provided in the headset: FP1, FP2, F7, F8, C3, C4, O1 and O2. Almost all of these electrodes can be integrated into the headband section of the headset, only C3 and C4 will be mounted on top of an arm extending to the upper half of the head. Ground and reference electrodes are placed just behind the ears. It is not entirely clear if we are going to need all of the eight available channels. Work has centred upon finding electrode positions which deliver the most efficient data, e.g. in respect with the relationships between pairs, or because of the robustness of the signal. Lowpass filters remove the airborne and omnipresent 50 Hz hum from electrical mains and other background noise that interferes with the reading of microvoltages on the scalp. Signal processing algorithms are written to treat the raw data and return scaled events to the Max/MSP computer which controls the pins on the steel segments.
Soon we will start analysing the brainwave content of these areas in order to map significant data into a format that is meaningful for the distribution of sound.

The international 10-20 system is based upon a division of the head from nasion (the point between the eyes on top of the nose) over cranium (top of scalp) to inion (the bump on the lower rear part of the skull). The “10-20″ refers to the percentages that delimit the electrode positions on this line that traverses the scalp.

Here is a more graphic representation:

electrode positions 10-20 frontal and lateral

20 channel EEG event

Brainwaves know different frequency ranges. Dynamic synchronic or asynchronic changes measured over a certain number of channels, can refer to certain events such as perception and recognition of images and objects, motoric intention or mental concentration.

The following frequency ranges are typical for EEG reading:

Alpha 7.5-13Hz Alpha waves are associated with relaxation. The amplitude of alpha waves ranges between 10 and 50 mV.
Beta 13-40Hz Beta waves are associated with alertness, arousal, problem solving, and concentration. Beta waves are fast but low amplitude.
Delta 0-4Hz Delta waves are associated with deep sleep. They are a high-amplitude, low-frequency wave, and are generated by the lack of processing by neurons. Delta waves can also be found when examining a comatose patient.
Theta 4-7Hz Theta waves are associated with sleep, but can also be associated with anxiety, epilepsy, and traumatic brain injury.
Mu 7-11Hz Mu rhythms are associated with the motor cortex, and can be used to recognise imaginary motor movement.

This artistic project is developed in a unique collaboration with a partner in the scientific field.
Research into the interpretation of brainwaves is provided by the world-leading research center in nanotechnology IMEC, based in Leuven, Belgium.

Their support is two-fold: building a prototype for the EEG-headset with IMEC’s expertise on wireless autonomous transducers and developing algorithms that are able to interpret the raw waveform data and to transform them in significant data which drive hammers and dampers on the steel plates.

www.imec.be