EEG with black people's hair in mind

Post Reply
Douwehorsthuis
Posts: 6
Joined: Thu Mar 15, 2018 6:06 pm

EEG with black people's hair in mind

Post by Douwehorsthuis »

Good morning,
I wanted to describe an issue we run into but for that I need to give some context:

It seems like EEG equipment was made using either mainly or only white people as a model. This potentially results in difficulties when we try to collect data from black people. We are in a unique and amazing place where our research is NOT made up from 99% white college students since we have a Lab situated in the Bronx. However, whenever people have braids, big afro's, oily hair products (even after taking multiple showers at home and at our lab), dreadlocks, etc. the quality of data goes down, or we need to send people home. To argue that this is a screening issue is saying that a group of people should not be involved science to begin with (not saying that anyone says this, but I wanted to prevent this argument).

Examples of specific problems:
1) dreadlocks block a lot of channels and result in the caps not fitting the way they should, however there is often a lot of scalp area that we can reach if we move the dreads around.
2) braids can block a lot of channels, we need to wiggle the cap electrodes a bit so that they fall between the braids. This result often in over-gelling noisy channels, slightly misplaced channels, and probably more problems.
3) Gel + Oily hair product that sometimes don't leave the hair completely after washing could increase the offset if I am correct, plus I assume it could have an effect on electrical impedance
4) Afro's make that the caps, aren't in the right place anymore. This means that the frontal channels move up, and sometimes there is no way they could reach the forehead. (I am talking about hair that is maybe 2+ cm high)


My question is this
: Is there specific bio-semi hardware/equipment that would help us or take these issues in consideration?

Thank you for taking the time to answer this.

Best,
Douwe Horsthuis

*I am using the term black people, since this seems to be the preferred way how black people, we spoke with, like to be addressed in this context. (Since not all black people that visit our lab are born in America and thus aren't African-American.)

mlwoody
Posts: 1
Joined: Wed Aug 26, 2020 6:07 pm

Re: EEG with black people's hair in mind

Post by mlwoody »

Thank you for posting this question - I came to this forum to ask a similar one. I recently came across a preprint entitled "Novel Electrodes for Reliable EEG Recordings on Coarse and Curly Hair", which detailed a novel approach for addressing many of the barriers you describe. More info here: https://www.biorxiv.org/content/10.1101 ... 6.965202v1 I am curious to know how/if BioSemi is innovating in this area, as progress here is essential for diversity and inclusion in research. I will follow up with the BioSemi team directly but am also eager to know how others using BioSemi hardware are moving forward to better serve impacted BIPOC stakeholders.

Best wishes,
Mary Woody, PhD
Assistant Professor of Psychiatry
University of Pittsburgh

Coen
Site Admin
Posts: 991
Joined: Fri Mar 26, 2004 7:00 pm
Location: Amsterdam, Netherlands
Contact:

Re: EEG with black people's hair in mind

Post by Coen »

The BioSemi system uses innovative active electrodes. The active electrodes tolerate high electrode impednaces (up to hundreds kiloOhm). So long as the electrode offsets are within the acceptable range (say between +100 mV and - 100 mV) and stable, clean EEG signals can be acquired. Skin preparation (scrubbing) to lower impedances is not required and not recommended.

For subjects with thick/dense/curly hair, we recommend to use headcaps with the smallest 11 mm diameter electrode holders, instead of the 14 or 17 mm diameter holders that are standard on 32 and 64 channel headcaps.

Best regards, Coen (Biosemi)

Douwehorsthuis
Posts: 6
Joined: Thu Mar 15, 2018 6:06 pm

Re: EEG with black people's hair in mind

Post by Douwehorsthuis »

The suggestion of using smaller electrode channels is not new. Indeed, it seems to be the solution BioSemi agreed to put forward as the working solution.
Though I am confident that this solution is science-based and that you have tested this in different hair types, I am left wondering how this really solves the issues I was hoping to get addressed given that with certain hair types: one will still need to add significant more gel; the cap will still be further from the scalp; the cap may be too loose, etc., all contributing to a reduction of accuracy.
Interesting how we seem to have different data quality standards for distinct racial and ethic groups.
As scientists, we must work towards diversity and equality in science. This becomes difficult when scientists and the companies that develop scientific equipment have such different goals.

Douwehorsthuis
Posts: 6
Joined: Thu Mar 15, 2018 6:06 pm

Re: EEG with black people's hair in mind

Post by Douwehorsthuis »

Dear Mary Woody,
Thank you for replying and being vocal about this issue. I think the thought of irrelevance for the need for change or solutions stems in part from the lack of sound from people that problem solve on their own instead of as a group. I have seen the preprint before and even though it is great that people are genuinely trying to solve this problem, I've seen so many hairstyles that would still be exclude. It is also very difficult for people with sensory issues to go through un-braiding/braiding their hair.

As for serving the BIPOC stakeholder this is what I try, and since this is a sensitive topic I try my best to use the correct wording, but if someone sees something that I should call differently please let me know.

This is what I try to do so I don't need to exclude anyone from a study,
when people have braids (dreadlocks are similar but it makes it often more likely to loose more channels) I try to wiggle the electrode holders so that these fall in between the braids and thus are as close as possible to the scalp. The upside of braids is that they are often in a shape that works pretty well with a 20-20 system and a 64channel cap. The downside is that this does alter slightly the location of some electrods, some electrodes end up being at an agle (not sure if that matters) and if the braids are too close together this might not work at all.

When people have an afro (first of if it's more than a cm and a half or so I am not certain of the quality of the data and I would love for someone to look into this) I try to find a cap that still has the FP(1,z,2) channels on the forehead without being too loose on the sides. After that when adding gel, I unfortunately see no other way then adding significantly more gel compared to people with different hairstyles (again I would love for someone to let me know a better solution or tell me how problematic this is). I use Signa Gel, and I have heard that using cream instead might be better (less viscosity, but I don't know if there is a downside to is).

Are there any other people that have other solutions, different problems, ideas or answers?

Post Reply