IEEE Consumer Electronics Magazine - July 2016 - 58

standards under development within the CTA R6 SC4 WG3
committee. The more populated the eeG databank (in the
context of big data), the more reliable the data that one can
retrieve to allow a statistically rich pattern matching for recognition of a brain state. Simple vocabularies like the four suggested above can be expanded once sharable eeG data
formats are adopted. it may behoove the industry to come up
with a standard portable eeG signal collection tool to assure
eeG is uploaded in a synchronized and harmonized format.

DEBATES ABOUT BRAIN STATES
it is easy to challenge the methodology and vocabularies suggested in Figure 2. Plenty of models exist in neuropsychology
and beyond. A few descriptions of brain states have been classified through tests and the development of systems drawn from
not just the brain but also face and speech recognition [5]-[9]:
▼ Russell's model: surprise, excitement, happiness, calmness, sleepiness, depression, sadness, and stressfulness
▼ Facial Action Coding System (FACS): anger, disgust, fear,
joy, sadness, and surprise
▼ Sophisticated Highspeed Object Recognition engine
(SHORe): happy, sad, angry, surprised, and neutral; also
in style of cheerful, depressed, aggressive, astonished, etc.
Social media captures emotions from the web, cameras,
and the upcoming new game interactive movies, which
record emotional input by using agree or disagree, like versus
dislike, thumbs up and down, and many more examples triggered by a voluntary user's input from a keyboard or other
command devices, without eeG signals. And the fun, simple
graphic emoji [16] is yet another tool to show prevalent emotions by users themselves.
Said vocabularies describing brain states can stir endless
debates among physicians, psychiatrists, and humanists, let
alone scientists and engineers who attempt to measure the
narrow-channel eeG signals to portrait a user's brain state.
Questions remain: how should the state of the brain be
described? And, by the author, how many states can be
described by eeG data alone?
elaborate algorithmic research to harvest the weak eeG
signal continues making progress on its own pace. inconsistent eeG data continue to be collected for attention and meditation, using consumer headsets in nonreal time fashion [14],
[15], [18]. Little public access is accommodated. Some
examples have combined brain waves (bands: a, b, D, c, n,
and i) from headsets and revealed in their respective marketing brochures, summarized verbatim as follows:
▼ D (1-4 Hz) oscillations-in high emotional conditions or
in a sleep stage...dreaming
▼ i (4-7 Hz) oscillations-similar to Δ wave, calm,
relaxed...sleep
▼ a (8-12 Hz) oscillations-smooth patterns, awake calm,
eyes closed relaxed...alertness
▼ n (8-13 Hz) oscillations-from sensorimotor cortex...
moving hands and feet
▼ b (12-30 Hz) oscillations-desynchronized, normal awake
open eyes, busy, churning, concentrating...rest
58 IEEE Consumer Electronics Magazine

JULY 2016

(25-100 Hz) oscillations-from somatosensory cortex -
c
busy, churning, concentrating....
The above heuristics do not make the description of a clear
brain state easy. The brain state is not solely reflected by any
single band with discrete frequencies. is it possible to depict a
consumer's brain state simply and exclusively after all? if in
doubt, then eeG test results should be preserved in raw form
to allow further proof of eeG's validity in describing the
state of brain.
On the other hand, choice of states should be kept simple
and left to the application's manipulation, even for a rehabilitation special effect. One would hope that applications
would not be constrained by the emotion subtlety between
calm and relaxed and simply look for a distinctive change
to reflect the opposite mood. Therefore, states that indicate
positive or negative associated with a priori state can be
more feasibly observed with a clear difference in their
respective eeG profile. This binary-mode approach could
make the data design, storage, and retrieval of eeG signals
doable for any state, while the state would be associated
with a collection of eeG signals manifested in various patterns to share across the internet carrying legitimate metadata. it is expected that efforts devoted to specifications
#2057, #2058, #2060, and #2061 would benefit and lay the
cornerstones for #2059 to provide useful, consistent, and
composite brain data under stimulus in the first place. it is
also clear that, without #2059, standardization would be
boxed in, unable to contribute to commercial connected
applications readily.
Therefore, it is recommended that a state of the brain be
defined by individual applications. The description of a brain
state can carry a label of P, N, S, or W for positive, negative,
sick, and well brain state, respectively. Following the format
structure to be described in the future specification #2059,
one can establish a databank, in the context of big data, processing eeG databases from various collection sources.

▼

OTHER NETWORKING CONSIDERATIONS
Other labels of (2), (3), and (4) in Figure 1 are explained as follows: Label (2) promotes big data analytics for eeG pattern
search to be used in near real-time processing feasible for
dynamic recognition of brain state (change). Label (3) refrains
use of privacy information in eeG data manipulation. User private information should be removed from the eeG databank in
the cloud. Metadata structure is under review for specification
#2060, and Qusp has proposed XDF as a data storage format
[17]. Label (4) allows application events to be triggered by
mechanisms via computer keyboards, wheel chair control buttons, game consoles, smart phones, smart watches, 3-D glasses,
etc. These application enhancements are left to the innovative
developers to choose outside of the scope of standardization.

REQUIREMENTS FOR SPECIFICATION #2059
ON USER STATE DESCRIPTION
Brain states are difficult to be described in common, unambiguous vocabularies. An alternative is to associate each brain state

Table of Contents for the Digital Edition of IEEE Consumer Electronics Magazine - July 2016