It is difficult to place electrodes, then electrodes get tilted with time during session. Then the signal looks bit noisy as comparatively.
ImpactNoisy, degraded signal quality
Solution would help
Artefacts need to be filtered out. Noise signals affect efficiency of data processing.
ImpactAffects efficiency of data processing
Solution would help
Certain types of curly hair, afro hair styles, and braided hair are often difficult to work with, due to electrode placement and gel. Typically it requires more patience and willingness to give up on certain electrodes.
ImpactOften requiring electrode dropping due to low impedance or loose placement
Solution would help
Poor impedance and significant noise in the data. EEG caps are not designed to accommodate all hair types and depend on thin hair types.
ImpactHave to exclude data points because of noisy data
Solution would help
Difficult to get good contact with scalp or impedances. It is difficult to make good contact with the scalp.
ImpactLess diverse representation of populations with coarse or curly hair
Solution would help
Impedance levels stay high, very high S/N ratio, impedance is flaky.
ImpactSome data has to be excluded; pseudo-randomisation at worst
Solution would help
Had to put a lot more gel to get to the scalp. There are gaps between the chip and scalp in the electrodes.
ImpactExtra time for data collection and clean up
Solution would help
Difficulties getting readings; need to use more gel than with Eastern or Caucasian hair; hairstyles like braids prevent the cap fitting well. EEG caps are designed for participants with Caucasian hair.
ImpactHave to discard a lot of data; samples become less representative and diverse
Solution would help
High follicle density means it's harder to get wet electrode contact with the scalp. Unable to part hair successfully.
ImpactLow SNR signals; harder to extract resulting effects
Solution would help
In fact it was smoother than regular straight hair and data quality was good. I did not face any such challenges.
ImpactNo challenges reported in this instance
Solution would help
Current EEGs aren't versatile enough for taking measurements for people with diverse hair types. Only a specific population of individuals can be utilised for studies.
ImpactOnly a specific population of individuals can be utilised for studies
Solution would help
Noisy signal not rectified by usual solutions, cap not fitting close enough to scalp, gel cleaning issues harder for individuals.
ImpactAffects generalisability, participant experience, and potential sample size
Solution would help
Low connectivity, high signal noise, gel can be hard to apply correctly. Compression of the hair pushing against the cap causes a large gap between the head and the electrode.
ImpactHair type becomes an explicit exclusion criterion
Solution would help
Even someone with loose kinky hair was an issue; it took an extra 30-40 min to cap, and we were not fully successful. A grad student shoved wooden sticks through the electrodes. The difficulty of capping actually meant that we scraped the scalp so hard there was a bit of blood.
ImpactBrain research on African American language perception made almost impossible; systematic bias beyond individual participants
Solution would help
Poorer electrode signal, way more noise, lower adherence to scalp.
ImpactNoisier results and a less diverse sample
Solution would help
I had weave and box braids throughout my EEG experiences and found researchers and clinicians struggling to place the electrodes and the cap. Cleaning was a mess — I still had gel in my hair days later and had trouble washing it out.
ImpactPoor participant experience from setup through to cleanup
Solution would help
Hair is usually too messy and it is difficult to place the EEG cap tensely on the subject's scalp. Curly hair tends to occupy much more space than straight hair.
ImpactElectrodes may not make enough connection, requiring much stricter preprocessing
Solution would help
Even for long or dense afro hairstyles, it works perfectly in 90% of cases. Only wigs or braided hair in a ponytail style do not work.
ImpactLimited cases; wigs and certain ponytail braids remain problematic
No solution needed
It's harder to move hair out of the way to get direct electrode-to-skin contact. Impedance is generally higher and fewer electrodes are fully prepped.
ImpactHigher impedance across the board; fewer electrodes reach full prep
Solution would help
People with coily hair generally have dedicated wash days. Headsets are not designed with these hair types in mind. Dry headsets that fit over afro-like hair would be much better.
ImpactNon-representative samples; skews results when different demographics excluded from data collection
Solution would help
High impedance and higher variability with dry electrodes. Weak electrode-skin connection and higher resistance.
ImpactPoor signal-to-noise ratio
Solution would help
Not handle.
ImpactResults affected
No solution needed
Generally we just exclude participants with this hair type because it complicates the research process, especially as we're not taught how to deal with these scenarios. It feels like that's a sentiment echoed by my colleagues across the world.
ImpactLack of diversity in datasets, fewer participants, longer setup with often poorer signals
Solution would help
Length is the biggest issue. But curlier hair is easier to part when using gel and syringe because once parted it's more likely to stay.
ImpactPartitioning technique key; curlier hair manageable with right approach
Solution would help
Caps from non-EGI systems lift electrodes off scalp. EGI has a tall sensor net that works well with dense and curly hair.
ImpactAbility to include people regardless of hair type is essential for many research studies
Solution would help
Gel sometimes can hardly touch the scalp for thick curly hair. Electrodes can hardly stay still on the scalp by themselves.
ImpactElectrode shift leads directly to EEG feature shift
Solution would help
Higher impedances, need more conductive gel, cap doesn't rest as cleanly on scalp.
ImpactResults more likely to be noisy; preprocessing may accidentally discard important signals
Solution would help
With cornrows it was simply not possible. We would have needed to damage the braiding in order to get contact.
ImpactPrep takes longer; with specific hairstyles testing becomes unfeasible entirely
Solution would help
The set-up time is longer as some electrodes struggle to make good contact, so you need to re-gel them. I rather not choose the voluminous hair participants.
ImpactChallenging at set-up stage; signal usually ends in normal range
No solution needed
Collected data was harder to classify compared to participants with non-curly or non-kinky hair. Achieved classification was extremely poor.
ImpactExcludes a significant number of potential participants, creating biased datasets
Solution would help
Cap too small, not fitting with my locks, gel not adapted, left build-up, electrodes kept disconnecting. I had to start gelling my head 10-15 min prior to the session. Itchy gel, electrodes entangled with hair.
ImpactPractitioner was cautious so no data impact, but the experience itself was poor throughout
Solution would help
If a full head silicone cap is used, this pulls the hairs out as it catches on coarse hair. I never had a Black participant return for a follow-up reading.
ImpactErroneous readings from discomfort; zero Black participant retention for follow-up studies
Solution would help
We don't admit anyone into a study who has braids because it is very difficult. If the signal quality isn't good enough, we toss the data — wasted experiment.
ImpactBlanket exclusion of braided hairstyles; wasted experiments when signal quality fails
Solution would help
Curly hair has not been a problem for me, however, I did not have any African phenotype participants with extreme density curls.
ImpactLonger prep times; no African phenotype participants with extreme density curls in sample to date
Solution would help
