Knowledge From the Afrikan World, For the World

PMC Articles

Developing Best Practices for Inclusion in fNIRS Research: Equity for Participants With Afro‐Textured Hair

PMCID: PMC12877426

PMID: 41645562

Abstract

ABSTRACT Functional near‐infrared spectroscopy (fNIRS) is a popular optical neuroimaging method; however, participants with Afro‐textured (i.e., dark, coarse, curly) hair are often excluded due to difficulty obtaining sensor–scalp contact. Grounded in lived experience and sociocultural literature, we aimed to develop and evaluate culturally responsive best practices for participant interaction and hair preparation to increase Black participant inclusion in fNIRS research. First, we developed an intake survey, guidelines for researcher staffing and training, and a suite of customizable hair preparation techniques that prioritize participant comfort and hair integrity. We then evaluated these techniques with 19 Black participants (11 adults, eight children) with varying hair types/styles; methods included braiding cornrows around the intended optode montage, using gels and clips to part hair, and various ways of increasing tension to promote sensor–scalp contact. On average, signal quality improved by 50%, with the greatest improvements in anterior regions. While signal quality was not perfect, it was generally improved to the point of acceptability and inclusion in a racially/ethnically diverse dataset (with hair type/color as covariates). We conclude with recommendations for increasing awareness of racial bias in neuroimaging, greater diversity in research teams, and a more inclusive approach for working with diverse populations.

Full Text

Functional near‐infrared spectroscopy (fNIRS) is an optical, mobile, and motion‐tolerant neuroimaging method well‐suited for brain analyses in naturalistic, real‐world settings, thus providing wide opportunities to collect ecologically valid neuroimaging data. However, the optical properties of near‐infrared light and the design of fNIRS hardware are often incompatible with hair of darker color and coilier texture (e.g., Afro‐textured hair). Accordingly, many fNIRS research findings have been based on samples that exclude Black participants with Afro‐textured hair (Kwasa et al. 2023). Thus, the present study aimed to develop a series of culturally responsive, participant‐centered best practices for fNIRS participants (adults and children) with Afro‐textured hair and to empirically evaluate whether these procedures improve signal quality and, ultimately, participant inclusion.
fNIRS is a noninvasive, cap‐based, optical neuroimaging technique utilized to index brain activation via the hemodynamic response in cortical regions of the brain. fNIRS sources shine harmless near‐infrared light into the skull that penetrates the outer layers of the brain's cortex a few centimeters deep. This near‐infrared light then scatters and reflects, and strategically placed detectors (approximately 3 cm from the sources) measure the light that is reflected back out to the scalp through a “banana” shaped path. Because oxygenated and deoxygenated blood have different hemodynamic properties, and thus reflect different amounts of light, these features allow estimation of the intensity of brain activity taking place in specific measured regions of interest (Ferrari and Quaresima 2012; Quaresima and Ferrari 2019; Scholkmann et al. 2014).
Scholars have increasingly employed fNIRS as a tool over the past decade because of its many advantages, including lower cost, portability, and motion tolerance (relative to other techniques such as MRI and EEG). The portability of fNIRS invites the possibility of data collection in more naturalistic and ecologically valid non‐lab settings, which can increase the inclusion of participants facing barriers to participating in lab‐based research at a university or medical center (e.g., transportation issues, dislike of medical settings) (Arredondo 2021; Pinti et al. 2018). Additionally, the relative motion tolerance is well‐suited for individuals or populations who might struggle to sit perfectly still for long periods of time, such as young children and those with developmental or psychiatric disabilities (Blasi et al. 2019; Rahman et al. 2020; Wilcox and Bioindi 2015; Zhang and Roeyers 2019). However, research with fNIRS is not without technological limitations, as discussed below.
“Afro‐textured” is a term used to describe features of the naturally curly hair common in people of African descent (henceforth referred to as Black). While people from all races can have curly hair, Afro‐textured hair is characterized by several features, including “type, texture, elasticity, porosity, density, and natural curl and shape” (Brown and Lemi 2021). Hair texture refers to the thickness of the hair strand; elasticity refers to the breakability of the hair; porosity refers to the hair's absorbency; density refers to number of strands per unit of scalp; and curl and shape refer to how bent or coiled each curl is, which can vary substantially across a single individual's scalp (Brown and Lemi 2021). While there are many classification systems for hair type, the most commonly used one in the Black hair community is the Andre Walker hair typing system (Figure 1), which groups hair types into one of four categories: Type 1, or straight hair; Type 2, or wavy hair; Type 3, or curly hair; and Type 4, or kinky hair (Divina Blk 2021; Trebilcock 2017). The majority of Black individuals have Type 3 or 4 hair, where Type 3 is described as having a “loopy” pattern in the shape of an “S,” where curls are defined and may have substantial volume, and Type 4 is described as visibly patterned with very tight curls (Ellis‐Hervey et al. 2016). While this hair‐typing system is not without criticism (e.g., Gaines et al. 2023), utilizing such a typology can be a helpful starting point for understanding features of Afro‐textured hair and the need for specialized approaches in neuroscience. Given the simplicity and familiarity of this system in the Black community, we use this system in communicating about hair type with participants before participating in neuroscience research.
Andre Walker hair typing system (image from Women Health Info Blog [2017]).
Because the fNIRS signal is light based, the optical properties of the hair, skin, and skull attenuate the signal quality (Figure 2). For individuals with fine, straight, light‐colored hair and lighter skin tones, it is relatively easy to achieve scalp–optode contact and, in most cases, usable signal quality. However, the data acquisition process is not as straightforward for participants of all hair and skin types.
Illustration of fNIRS signal loss across different hair types, hair colors, and skin pigments. Reprinted with permission from Kwasa et al. (2023).
These physical realities and technological biases have led to systemic exclusion of participants with Afro‐textured hair in fNIRS and other neuroscientific research.1 However, the rapidly growing body of fNIRS research exhibits extremely limited reporting of demographic factors and other relevant contextual information about participants, as noted by two recent systematic reviews. Girolamo and colleagues reviewed 38 fNIRS studies and found that only 5% reported participant race and ethnicity, compared to 89% that reported gender and 100% that reported geographic location (Girolamo et al. 2022). Similarly, Kwasa and colleagues evaluated 87 fNIRS papers from 2017 to 2022 and found that a mere 2% reported race and ethnicity, 0% reported skin tone, and 3% reported hair type, compared to over 90% that reported participant gender (Kwasa et al. 2023). Given this systematic underreporting, it is difficult to fully quantify the level of exclusion based on race, ethnicity, skin tone, and hair type and texture in fNIRS research, though it is likely similar if not even greater than that seen with other neuroscientific modalities (Bradford et al. 2024; Choy et al. 2021; Etienne et al. 2020; Kwasa 2024; Louis et al. 2022; Parker and Ricard 2022).
Thus, there is a critical and immediate need for improving fNIRS research approaches to increase Black participant inclusion. While technological innovation is certainly needed (see Yücel et al. 2024), we instead focus on developing approaches that researchers can employ using existing fNIRS technology. However, it is essential to first consider several cultural realities in participant‐centered best practices.
Due to Eurocentric beauty standards and persistent anti‐Black racism, Afro‐textured hair faces significant stigmatization. “Texturism” is the hierarchical positioning of straight hair and looser curl patterns above tighter, kinkier ones (Dixon and Telles 2017; Rowe 2021; Shepherd 2018). “Good hair” is often characterized by straight or wavy textures that align with Eurocentric standards, while “bad hair” is labeled as tightly coiled and thick (Robinson 2011). Many Black individuals choose to chemically relax their hair to appear straight, despite the potential damage this may cause, because “natural” hair texture can be a barrier to social and professional acceptance (Ellis‐Hervey et al. 2016; Gray 2017; Khumalo et al. 2010; Yosso 2005). This stigmatization is so pervasive that in 23 US states, it remains legal to discriminate against Black individuals for employment and educational purposes based on their hair texture and style (CROWN Coalition and Dove 2025). Because of this reality, many Black individuals consider that wearing their hair naturally is a political statement that represents Black pride, despite ongoing media portrayals favoring Eurocentric hair standards.
Relatedly, many individuals with Afro‐textured hair wear their hair in protective hairstyles, such as cornrows, braids, twists, locs, Bantu knots, weaves, and wigs. These styles prevent breakage and promote hair health, and may include only the wearer's own hair or may also incorporate extensions or other additional hairpieces (Sandeen and Hancock 2024). These styles often require considerable time (a full day or more) and expense (up to several hundred dollars) to implement, leading many to wear them for weeks, months, or even years.
Thus, hair is deeply tied to identity and culture within the Black community, so researchers should respect participants’ styling choices and never ask individuals to relax their hair or remove protective styles solely for research participation (Thompson 2008). Such requests are likely to be seen as ignorant and offensive, and reduce chances of inclusion.
In contrast, for those with Afro‐textured hair, “wash day” refers to a scheduled hair‐washing day, because this process can take several hours or even days to complete, depending on the amount and type of hair, whether a style is being removed, and the type of new style being implemented. The wash day process often involves detangling, washing, deep conditioning, and styling, which needs to be conducted with hair care products designed specifically for Afro‐textured hair to avoid damage to the hair follicle. To learn more about wash day, we recommend the monograph Wash Day: Passing on the Legacy, Rituals, and Love of Natural Hair by Tomesha Faxio (Faxio 2023, 2024; White‐Grier 2024).
Preparation for fNIRS studies almost always requires manipulation and “messing up” of a participant's hair. While this may be a minor inconvenience for many participants, for those with Afro‐textured hair, these manipulations need to be planned in advance, so that participants know what to expect and can plan enough time to return their hair to the way they want it. Ideally, researchers can schedule fNIRS sessions either immediately before or after participants’ intended wash day, when they were already planning on taking out the existing hairstyles. Additionally, researchers should prepare Afro‐textured hair with products specifically designed for Afro‐textured hair that will not cause long‐term damage to the hair (see Supporting Information S1 for a suggested shopping list).
A final consideration is the role of the hair salon or barbershop as something of a sacred place. These establishments function as places where community members can discuss challenging topics while celebrating relationships, resilience, and joy (Palmer et al. 2022; Solomon et al. 2004). Importantly, the loyalty between clients and hair care professionals runs extraordinarily deep, with relationships often lasting decades and extending beyond professional services. Stylists and barbers frequently become integrated into clients’ families, attending important events and sometimes even styling deceased clients’ hair for funerals. Given the deep trust and loyalty inherent in hair care relationships, participants with Afro‐textured hair may be hesitant to allow unknown researchers to manipulate their hair (Palmer et al. 2022). Combined with the history of stigmatization, participants may be understandably nervous, distrustful, and even alienated by well‐meaning but poorly informed researchers. This underscores the importance of developing culturally responsive approaches for fNIRS research and ideally, as described below, the inclusion of research team members who share similar values and positionality.
Promisingly, many scholars have recently published calls to action and specific recommendations to increase diversity, equity, and inclusion in human neuroscience, especially in EEG, which uses a cap and scalp‐based sensors similar to fNIRS (Adams et al. 2024; Etienne et al. 2020; Garcini et al. 2022; La Scala et al. 2023; Margolis et al. 2025; Mlandu et al. 2024; Murray et al. 2021; Richardson 2021; Webb et al. 2022; Wu et al. 2024). For example, Lietsel Richardson's “A Guide to Hair Preparation for EEG Studies” includes specific subject preparation recommendations, including style suggestions (braided crown, tight low bun, pigtail braids), and scheduling considerations, such as scheduling data collection to coincide with washday for removable styles or right before a retwist session for nonremovable styles (Richardson 2021). This may be especially relevant in pediatric populations, when investigators commonly reach out to families close to a child's birthday when they age into a new condition or project (Adams et al. 2024). However, birthdays and other major life events often coincide with special, and sometimes expensive, hairstyles being put in place (“birthday hair”) (Adams et al. 2024). These considerations underscore the importance of communication between the research team and research participants to maximize comfort and uptake.
Inclusive practices specifically for fNIRS research—and especially pediatric fNIRS—are much more limited. Although some researchers have recently discussed these and other equity‐related concerns with fNIRS (e.g., Cheaye and Rosen n.d.; Doherty et al. 2023; Eng 2024; Khan et al. 2012; Yücel et al. 2024), solutions have largely focused on hardware re‐design rather than specific best practices in participant interaction and hair preparation. Thus, we set out to develop participant‐centered best practices to improve inclusion of participants with Afro‐textured hair in fNIRS research. Rather than providing a simple, universally applicable solution, our work over the past several years has highlighted the importance of personalization and respect for the individual.
Nineteen individuals with Afro‐textured hair (11 adults, eight children; eight male, 11 female) participated in the present study. Child participants (ages 3–5 years) were participating in a longitudinal study of cognitive and brain development across preschool ages, and thus, the present investigation took place as a part of their planned fNIRS session in which they completed touchscreen‐based tasks of language, theory of mind, and executive functioning skills. Adult participants were friends, family, and colleagues of author A.S. who were specifically invited to participate in this investigation. All participants identified as Black/African American with Afro‐textured hair of different styles. Table 1 details participants’ sex, hair type/texture, porosity, wash frequency, and day‐of‐visit hairstyle. Unfortunately, signal quality data for Participant C1 were lost and thus not included in analyses, but we leave their information in the table for completeness. All participants (or their parent/guardian) provided informed consent and were compensated $20/h.
Data were acquired using a NIRx NIRSport2 mobile fNIRS system with 32 dual‐tip LED optodes (16 sources, 16 detectors). The montage was designed to cover brain regions associated with language, executive functioning, and social cognition, which includes bilateral prefrontal, superior/middle temporal, and temporoparietal regions (Figure 3, left). Ultimately, this created 44 topographic data channels of interest between adjacent sources and detectors (22 per hemisphere). Optodes were populated into the appropriately sized EasyCap, ranging in size from 52 cm (child participants) to 60 cm (adult male). The initial optode population used spring‐loaded tension level 1 on the forehead (anterior four optodes on each side), tension level 2 on the front/middle of the head (the next six optodes), and tension level 3 on the back of the head (posterior six optodes); this configuration was often altered (see below).
For 15 participants (including all seven children and eight adults), one detector (F4, in right superior frontal) was sacrificed and replaced with eight short‐distance detectors that were distributed evenly throughout the montage, whose purpose is to collect nonneural physiological signals for regressing out of the data from the primary channels. This left 42 long‐distance data channels, plus eight short‐distance data channels (Figure 3, right). Because there was less variation in the short‐distance channels, we primarily examined the improvement of the long‐distance data channels (see Supporting Information S2 and Figure S1 for short‐distance channels).
During the calibration process, the NIRSport2 device increases the brightness of each source in a stepwise manner (from 0% to 100%) until the optimal signal amplitude for each channel is obtained. The Aurora software then provides a stop‐light–colored rating system for each data channel to signal researchers which channels to improve before starting the experiment (Figure 4). Red indicates critical or poor signal quality; yellow indicates acceptable (but potentially improvable) signal quality; and green indicates excellent signal quality that needs no further improvement. The colored values are based on the raw voltage measured at each detector (red <0.5 mV; yellow = between 0.5 and 3 mV; green >3 mV). While these quantitative values are provided in real time during signal optimization, they change rapidly based on the physiological signal, while the coarser color categories are more stable and are typically the basis of researchers’ decisions to proceed. As stated in the Aurora user guide, “it is OK to proceed to recording with some yellow channels, but not advisable to proceed with red channels, as these will likely need to be excluded in later analysis for having a high coefficient of variation (C.V.), or noise” (NIRx Medical Technologies 2023, 25). Thus, we use the color categories as the primary outcome of interest.
Below, we describe our procedures in three sections: activities before the visit, during the visit, and after the visit. We also provide two flowcharts to aid in understanding these procedures: Figure 5 describes the recruitment and scheduling process, while Figure 6 describes the explicit hair preparation options during signal optimization.
Based on the reviewed literature, author A.S.’s lived experience, and our team's expertise in working with children and families, we developed an intake form that we refer to as our “hair equity screener” to send to participants (or pediatric participants’ caregivers) who would be engaging in fNIRS studies that require whole‐head capping (i.e., this may not be necessary if studies only use prefrontal montages that only touch the forehead). Note that the present version of this screener is designed for a study of preschool children aged 3–5 years (the full text of the screener is available in Supporting Information S3). It begins with a personal introduction to the Hair Equity Specialist to help establish trust. Alongside a photo of herself, she describes her Black identity and her experience styling her own natural hair since she was 13 years old. She continues, “I hope that seeing someone who looks like you will make you more comfortable with your child being in our study. Below are more specific questions on how you take care of your child's hair.” Anecdotal participant feedback expressed gratitude for the humanizing personal details about the research team and indicated that these details have been incredibly helpful in establishing trust, providing the foundation for a relationship between participants and researchers, and offering families background information that can help ease pediatric participant anxiety through preparation before the visits.
Next, we asked about specific products that the caregiver uses in the child's hair, including shampoo, conditioner, and any styling products (e.g., gel, cream), as well as whether or not they ever use heat (e.g., blow‐dryer). We collected information about curl type and texture and hair porosity by linking participants to an online quiz and descriptions to help them determine their types (Carol's Daughter 2023; Viola n.d.). Finally, we provided an open‐field response for anything else the caregiver may want to share about their child's hair, routine, personality, or other factors. In total, this intake form establishes trust between the participant and the research team while also allowing the researchers to adequately prepare for the participant.
If possible, participants were scheduled to come on or just before a wash day, when their hair was planned to be natural and loose. However, five adult participants were purposely invited to come while they had a protective style in place, in order to investigate possibilities for working around the style (see Table 1). After a participant completed the intake form, we scheduled the participant's lab visit. Notably, these visits typically took longer than visits for participants who do not have Afro‐textured hair. While the exact timing depends on the specific montage being used, the participant population, and their tolerance for sitting still (e.g., children, individuals with disabilities), we planned for an extra hour and compensated participants for the full time they spent in the lab. In Supporting Information S4, we provide examples of infographics we provide to participants on what to expect during their lab visit.
We acquired a group of hair tools and products commonly used on Afro‐textured hair to assist with hair preparation, including hair pics, mini elastic hair ties, scrunchies, rat tail combs, wide tooth combs, metal or plastic alligator clips, hairdryer with attachment, hair gel, hair cream, leave‐in conditioner, spray bottle, barbicide, and disinfectant jar. In addition, we also used some materials specifically for fNIRS preparation, including washable chalk markers, light‐up crochet hooks, clamps, and foam mannequin heads. The use of these materials is detailed below. Further, in Supporting Information S1, we provide a “shopping list” of all materials we currently keep on hand, details of how and why we use the products, and suggestions for where to purchase these items.
Below, we describe a “menu” of techniques that can be employed to increase the fNIRS signal with Afro‐textured hair, ranging from least to most involved. These procedures were developed for use with NIRx NIRSport 2 hardware and software; however, they should be generally applicable to any fNIRS setup. In general, we aimed to use the least invasive process that achieves a good quality signal (see Figure 6 for a flowchart). Different participants required different approaches based on their hair type, their willingness and tolerance for hair styling, and the specifics of the fNIRS montage being used. Some participants required a trial of multiple approaches iteratively to achieve the best result. However, throughout the hair preparation procedure, the participant's comfort and hair integrity were prioritized, and the researcher engaged in an active assent process with each additional step or strategy attempted.
Interventions were applied to the whole head for child participants, as they were participating in an existing study. However, for all adult participants, we only applied the intervention to the right side of the head and kept the left side as is, so that each participant could serve as their own control. Further, the intentional inclusion of five adult participants with existing protective styles allowed us to test our full range of potential intervention options. The exact interventions used are detailed in Table 1.
Regardless of the hairstyle that the participant arrived wearing (natural, protective style, relaxed, or other), our first approach was always to test whether we could simply gently move the hair to allow the optodes to lie flat on the participant's scalp. The cap with pre‐populated optodes was fitted over the participant's head, and signal quality was monitored in real time to determine how interventions are affecting signal quality. Optodes or channels that did not have an excellent signal were flagged, and a researcher gently slid an LED‐tip crochet hook (see Supporting Information S1) into a nearby, unused slit in the cap and moved the hair out of the direct path of the optode to allow the optode tip to reach the scalp. Occasionally, researchers would also try pulling the hair through nearby slits/holes in the cap (as would be done with a highlighting cap) to help keep hair out of the way. This intervention involves minimal discomfort, as the overall process is minimally invasive and the tips of the LED crochet hooks are blunted plastic.
When a participant had longer hair, tying or pinning the hair sections often did not sufficiently keep the volume of hair away from the parts. In this case, we parted the hair in the same way as Option 5a and then either cornrowed or flat‐twisted the hair sections as flat to the scalp as possible (Figure 7). Specifically, we started cornrows or twists near the optode‐dense regions and worked away toward regions without optodes, since the cornrows/twists became fatter as more hair was added and thus could push the optodes away from the scalp. While this was the most time‐intensive hair preparation option (it often took 30 min to an hour), it was also the most durable design, which may be preferred for longer studies where the fNIRS needs to be worn for a longer duration.
For one child participant who was extremely tender‐headed and apprehensive about having a stranger work with their hair, we instructed the participant's parent to create the study‐specific braiding at home or at a salon before coming for their research visit. While this limited the precision that could be achieved by the experienced researchers, it was deemed the only option to include this participant. The Hair Equity Specialist set up a video call with the child's parent to describe options, demonstrated where the parts would be on a mannequin head (Figure 8), described the measurements, and showed examples of prior participants (faces obscured). Ultimately, this resulted in partially usable data, as some braids were aligned with the montage, but some were slightly misaligned, such that the optode landed on top of a braid rather than a part. Researchers did not attempt to rebraid the hair.
Once the hair intervention was completed, we ran another signal optimization and took a screenshot. If additional interventions were needed, we repeated the process and took a final optimization screenshot after all interventions were complete. The total length of the hair intervention ranged from 30 min to 3 h. We also took photographs of the participant wearing the cap and additionally took a 3D whole head image using a Structure Sensor Pro (Occipital 2022) to use for post hoc evaluation of optode locations. Lastly, we removed the cap and took a final set of photographs (front/back, sides, and aerial) to document the final hair equity intervention procedure for recordkeeping. In addition to these photographs, we also kept written records of the participants’ skin color, hair color, hair texture, and hair style.
From preintervention to postintervention (Figure 9, top), there was a significant increase in the number of green channels (t(12) = 4.58, p < 0.001; mean increase = 5.85 channels) and a significant decrease in the number of red channels (t(12) = −5.79, p = 0.001; mean decrease = 5.69 channels). There was no significant change in the number of yellow channels. Among participants who had intervention applied to only half of the head (Figure 9, bottom), the intervened side (vs. non‐intervened side) had significantly more green channels (W = 0.0, p = 0.014; mean difference = 5.36 channels) and significantly fewer red channels (W = 55, p = 0.006; mean difference = 6.45 channels). There was no significant difference in the yellow channels between the intervened and non‐intervened sides. All four significant results survived FDR correction for multiple comparisons (all adjusted p < 0.03).
To facilitate a qualitative comparison across individual participants, we calculated the percent improvement in usable channels (green or yellow; see Table 1). From pre‐ to postintervention, improvement in usable channels ranged from 0% to 150%, with an average of 50%. Comparing right to left, improvement in usable channels ranged from −5% to 59% (note that the participant with fewer usable channels on the right/intervened side used the montage with two fewer data channels on the right). Although the sample is not large enough for formal subgroup analysis, it appears that signal improvement was greatest for participants who came with loose hair that was cornrowed, and improvement was minimal for participants with protective styles (e.g., box braids and locs). Additionally, while it appears that improvement was greater for adults versus children, children tended to have better baseline signal quality before intervention, leaving less room for improvement. Finally, visual inspection suggests that the greatest signal improvement was seen in more anterior regions, as compared to more posterior regions, where hair is often thicker.
Building on our experiences and findings, we offer several broader recommendations for researchers conducting fNIRS studies with participants who have Afro‐textured hair. First, whenever possible, research teams should hire and meaningfully involve staff with lived experience of having, caring for, and styling Afro‐textured hair. Such individuals may be students, community members, or professional stylists, and their involvement can be critical for establishing trust with participants, reducing anxiety, and minimizing unintentional harm to participants’ hair, comfort, and sense of dignity. Having individuals on research teams who hold similar identities to their participants can help ensure participants feel as comfortable as possible and may also allow participants to voice concerns with diminished fear of dismissal due to racial bias (Nicolaidis et al. 2010). Although non‐Black researchers can learn hair‐equity techniques, shared lived experience often facilitates communication and rapport in ways that technical training alone cannot. Importantly, whoever fills this role should be compensated appropriately for their time, expertise, and labor, with compensation structures that may reasonably differ depending on whether this individual is a student research assistant, a community consultant, or a professional stylist.
Though tool improvement has not been the focus of this investigation, we offer a few recommendations and ideas. First, the fNIRS cap design could better accommodate voluminous hair. For example, larger holes distributed throughout the cap could allow hair to be pulled through more effectively, enabling the cap to lie closer to the scalp. The optimal placement of such holes may vary depending on the study‐specific optode montage, suggesting that customizable cap designs would be ideal. Additionally, given the volume and density of Afro‐textured hair, it may be necessary to produce larger cap sizes overall. For instance, one commonly used brand, EasyCap (Brain Products), currently only extends to 64 cm in circumference, which may still be too restrictive for adults with very voluminous hair. Brush‐fiber–tipped optodes (e.g., Khan et al. 2012) may also help comb through dense or coily hair, and higher tension levels of spring‐loaded grommets could further improve scalp contact. However, we caution manufacturers to prioritize participant comfort to avoid introducing pain‐related inequities for Black participants.
Additionally, while we have focused primarily on signal attenuation due to hair, another potential source of bias in fNIRS data with Black participants is attenuation related to higher levels of melanin in the skin. Existing fNIRS technologies generally assume a fixed light pathlength through the brain across participants; however, because melanin is highly absorbent, individuals with darker, more melanated skin may experience systematic, nonlinear reductions in signal strength. This attenuation could lead to inaccurate estimations of relative changes in oxygenation in regions of interest (Kwasa et al. 2023; Wassenaar and van den Brand 2005). A full discussion of melanin‐related limitations and methodological solutions is beyond the scope of this paper, but readers are directed to prior and emerging work in this area (e.g., Bronkhorst et al. 2019; Patel et al. 2024; Roy et al. 2024; Sardar et al. 2001; Yücel et al. 2024).
In conclusion, this investigation aimed to develop and evaluate best practices for conducting fNIRS research with participants who have Afro‐textured hair. Without efforts at multiple system levels to diversify the research workforce, renounce scientific racism and exclusion, and repair damaged relationships with minoritized communities, calls to simply diversify samples fall short (Murray et al. 2021; Varma et al. 2021). As such, we hope that this investigation opens conversations about research practices at several levels of analysis.

Sections

"[{\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0028\"], \"section\": \"Introduction\", \"text\": \"Functional near\\u2010infrared spectroscopy (fNIRS) is an optical, mobile, and motion\\u2010tolerant neuroimaging method well\\u2010suited for brain analyses in naturalistic, real\\u2010world settings, thus providing wide opportunities to collect ecologically valid neuroimaging data. However, the optical properties of near\\u2010infrared light and the design of fNIRS hardware are often incompatible with hair of darker color and coilier texture (e.g., Afro\\u2010textured hair). Accordingly, many fNIRS research findings have been based on samples that exclude Black participants with Afro\\u2010textured hair (Kwasa et\\u00a0al. 2023). Thus, the present study aimed to develop a series of culturally responsive, participant\\u2010centered best practices for fNIRS participants (adults and children) with Afro\\u2010textured hair and to empirically evaluate whether these procedures improve signal quality and, ultimately, participant inclusion.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0019\", \"dev70134-bib-0040\", \"dev70134-bib-0049\"], \"section\": \"fNIRS as a Research Tool\", \"text\": \"fNIRS is a noninvasive, cap\\u2010based, optical neuroimaging technique utilized to index brain activation via the hemodynamic response in cortical regions of the brain. fNIRS sources shine harmless near\\u2010infrared light into the skull that penetrates the outer layers of the brain's cortex a few centimeters deep. This near\\u2010infrared light then scatters and reflects, and strategically placed detectors (approximately 3\\u00a0cm from the sources) measure the light that is reflected back out to the scalp through a \\u201cbanana\\u201d shaped path. Because oxygenated and deoxygenated blood have different hemodynamic properties, and thus reflect different amounts of light, these features allow estimation of the intensity of brain activity taking place in specific measured regions of interest (Ferrari and Quaresima 2012; Quaresima and Ferrari 2019; Scholkmann et\\u00a0al. 2014).\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0002\", \"dev70134-bib-0039\", \"dev70134-bib-0003\", \"dev70134-bib-0041\", \"dev70134-bib-0061\", \"dev70134-bib-0066\"], \"section\": \"fNIRS as a Research Tool\", \"text\": \"Scholars have increasingly employed fNIRS as a tool over the past decade because of its many advantages, including lower cost, portability, and motion tolerance (relative to other techniques such as MRI and EEG). The portability of fNIRS invites the possibility of data collection in more naturalistic and ecologically valid non\\u2010lab settings, which can increase the inclusion of participants facing barriers to participating in lab\\u2010based research at a university or medical center (e.g., transportation issues, dislike of medical settings) (Arredondo 2021; Pinti et\\u00a0al. 2018). Additionally, the relative motion tolerance is well\\u2010suited for individuals or populations who might struggle to sit perfectly still for long periods of time, such as young children and those with developmental or psychiatric disabilities (Blasi et\\u00a0al. 2019; Rahman et\\u00a0al. 2020; Wilcox and Bioindi 2015; Zhang and Roeyers 2019). However, research with fNIRS is not without technological limitations, as discussed below.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0006\", \"dev70134-bib-0006\", \"dev70134-fig-0001\", \"dev70134-bib-0011\", \"dev70134-bib-0055\", \"dev70134-bib-0014\", \"dev70134-bib-0020\"], \"section\": \"Features of Afro\\u2010Textured Hair\", \"text\": \"\\u201cAfro\\u2010textured\\u201d is a term used to describe features of the naturally curly hair common in people of African descent (henceforth referred to as Black). While people from all races can have curly hair, Afro\\u2010textured hair is characterized by several features, including \\u201ctype, texture, elasticity, porosity, density, and natural curl and shape\\u201d (Brown and Lemi 2021). Hair texture refers to the thickness of the hair strand; elasticity refers to the breakability of the hair; porosity refers to the hair's absorbency; density refers to number of strands per unit of scalp; and curl and shape refer to how bent or coiled each curl is, which can vary substantially across a single individual's scalp (Brown and Lemi 2021). While there are many classification systems for hair type, the most commonly used one in the Black hair community is the Andre Walker hair typing system (Figure\\u00a01), which groups hair types into one of four categories: Type 1, or straight hair; Type 2, or wavy hair; Type 3, or curly hair; and Type 4, or kinky hair (Divina Blk 2021; Trebilcock 2017). The majority of Black individuals have Type 3 or 4 hair, where Type 3 is described as having a \\u201cloopy\\u201d pattern in the shape of an \\u201cS,\\u201d where curls are defined and may have substantial volume, and Type 4 is described as visibly patterned with very tight curls (Ellis\\u2010Hervey et\\u00a0al. 2016). While this hair\\u2010typing system is not without criticism (e.g., Gaines et\\u00a0al. 2023), utilizing such a typology can be a helpful starting point for understanding features of Afro\\u2010textured hair and the need for specialized approaches in neuroscience. Given the simplicity and familiarity of this system in the Black community, we use this system in communicating about hair type with participants before participating in neuroscience research.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0062\"], \"section\": \"\", \"text\": \"Andre Walker hair typing system (image from Women Health Info Blog [2017]).\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-fig-0002\"], \"section\": \"Barriers to fNIRS Research for Participants With Afro\\u2010Textured Hair\", \"text\": \"Because the fNIRS signal is light based, the optical properties of the hair, skin, and skull attenuate the signal quality (Figure\\u00a02). For individuals with fine, straight, light\\u2010colored hair and lighter skin tones, it is relatively easy to achieve scalp\\u2013optode contact and, in most cases, usable signal quality. However, the data acquisition process is not as straightforward for participants of all hair and skin types.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0028\"], \"section\": \"\", \"text\": \"Illustration of fNIRS signal loss across different hair types, hair colors, and skin pigments. Reprinted with permission from Kwasa et\\u00a0al. (2023).\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-note-0001\", \"dev70134-bib-0022\", \"dev70134-bib-0028\", \"dev70134-bib-0004\", \"dev70134-bib-0009\", \"dev70134-bib-0016\", \"dev70134-bib-0027\", \"dev70134-bib-0030\", \"dev70134-bib-0037\"], \"section\": \"Barriers to fNIRS Research for Participants With Afro\\u2010Textured Hair\", \"text\": \"These physical realities and technological biases have led to systemic exclusion of participants with Afro\\u2010textured hair in fNIRS and other neuroscientific research.1 However, the rapidly growing body of fNIRS research exhibits extremely limited reporting of demographic factors and other relevant contextual information about participants, as noted by two recent systematic reviews. Girolamo and colleagues reviewed 38 fNIRS studies and found that only 5% reported participant race and ethnicity, compared to 89% that reported gender and 100% that reported geographic location (Girolamo et\\u00a0al. 2022). Similarly, Kwasa and colleagues evaluated 87 fNIRS papers from 2017 to 2022 and found that a mere 2% reported race and ethnicity, 0% reported skin tone, and 3% reported hair type, compared to over 90% that reported participant gender (Kwasa et\\u00a0al. 2023). Given this systematic underreporting, it is difficult to fully quantify the level of exclusion based on race, ethnicity, skin tone, and hair type and texture in fNIRS research, though it is likely similar if not even greater than that seen with other neuroscientific modalities (Bradford et\\u00a0al. 2024; Choy et\\u00a0al. 2021; Etienne et\\u00a0al. 2020; Kwasa 2024; Louis et\\u00a0al. 2022; Parker and Ricard 2022).\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0065\"], \"section\": \"Barriers to fNIRS Research for Participants With Afro\\u2010Textured Hair\", \"text\": \"Thus, there is a critical and immediate need for improving fNIRS research approaches to increase Black participant inclusion. While technological innovation is certainly needed (see Y\\u00fccel et\\u00a0al. 2024), we instead focus on developing approaches that researchers can employ using existing fNIRS technology. However, it is essential to first consider several cultural realities in participant\\u2010centered best practices.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0012\", \"dev70134-bib-0044\", \"dev70134-bib-0050\", \"dev70134-bib-0043\", \"dev70134-bib-0014\", \"dev70134-bib-0023\", \"dev70134-bib-0026\", \"dev70134-bib-0064\", \"dev70134-bib-0010\"], \"section\": \"Stigmatization of Afro\\u2010Textured Hair and Protective Styles\", \"text\": \"Due to Eurocentric beauty standards and persistent anti\\u2010Black racism, Afro\\u2010textured hair faces significant stigmatization. \\u201cTexturism\\u201d is the hierarchical positioning of straight hair and looser curl patterns above tighter, kinkier ones (Dixon and Telles 2017; Rowe 2021; Shepherd 2018). \\u201cGood hair\\u201d is often characterized by straight or wavy textures that align with Eurocentric standards, while \\u201cbad hair\\u201d is labeled as tightly coiled and thick (Robinson 2011). Many Black individuals choose to chemically relax their hair to appear straight, despite the potential damage this may cause, because \\u201cnatural\\u201d hair texture can be a barrier to social and professional acceptance (Ellis\\u2010Hervey et\\u00a0al. 2016; Gray 2017; Khumalo et\\u00a0al. 2010; Yosso 2005). This stigmatization is so pervasive that in 23 US states, it remains legal to discriminate against Black individuals for employment and educational purposes based on their hair texture and style (CROWN Coalition and Dove 2025). Because of this reality, many Black individuals consider that wearing their hair naturally is a political statement that represents Black pride, despite ongoing media portrayals favoring Eurocentric hair standards.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0046\"], \"section\": \"Stigmatization of Afro\\u2010Textured Hair and Protective Styles\", \"text\": \"Relatedly, many individuals with Afro\\u2010textured hair wear their hair in protective hairstyles, such as cornrows, braids, twists, locs, Bantu knots, weaves, and wigs. These styles prevent breakage and promote hair health, and may include only the wearer's own hair or may also incorporate extensions or other additional hairpieces (Sandeen and Hancock 2024). These styles often require considerable time (a full day or more) and expense (up to several hundred dollars) to implement, leading many to wear them for weeks, months, or even years.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0054\"], \"section\": \"Stigmatization of Afro\\u2010Textured Hair and Protective Styles\", \"text\": \"Thus, hair is deeply tied to identity and culture within the Black community, so researchers should respect participants\\u2019 styling choices and never ask individuals to relax their hair or remove protective styles solely for research participation (Thompson 2008). Such requests are likely to be seen as ignorant and offensive, and reduce chances of inclusion.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0017\", \"dev70134-bib-0018\", \"dev70134-bib-0060\"], \"section\": \"The Role of Wash Day and Using Black Hair Care Products\", \"text\": \"In contrast, for those with Afro\\u2010textured hair, \\u201cwash day\\u201d refers to a scheduled hair\\u2010washing day, because this process can take several hours or even days to complete, depending on the amount and type of hair, whether a style is being removed, and the type of new style being implemented. The wash day process often involves detangling, washing, deep conditioning, and styling, which needs to be conducted with hair care products designed specifically for Afro\\u2010textured hair to avoid damage to the hair follicle. To learn more about wash day, we recommend the monograph Wash Day: Passing on the Legacy, Rituals, and Love of Natural Hair by Tomesha Faxio (Faxio 2023, 2024; White\\u2010Grier 2024).\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-supinfo-0001\"], \"section\": \"The Role of Wash Day and Using Black Hair Care Products\", \"text\": \"Preparation for fNIRS studies almost always requires manipulation and \\u201cmessing up\\u201d of a participant's hair. While this may be a minor inconvenience for many participants, for those with Afro\\u2010textured hair, these manipulations need to be planned in advance, so that participants know what to expect and can plan enough time to return their hair to the way they want it. Ideally, researchers can schedule fNIRS sessions either immediately before or after participants\\u2019 intended wash day, when they were already planning on taking out the existing hairstyles. Additionally, researchers should prepare Afro\\u2010textured hair with products specifically designed for Afro\\u2010textured hair that will not cause long\\u2010term damage to the hair (see Supporting Information S1 for a suggested shopping list).\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0036\", \"dev70134-bib-0052\", \"dev70134-bib-0036\"], \"section\": \"Cultural Significance of the Hair Salon and Relationships With Barbers/Beauticians\", \"text\": \"A final consideration is the role of the hair salon or barbershop as something of a sacred place. These establishments function as places where community members can discuss challenging topics while celebrating relationships, resilience, and joy (Palmer et\\u00a0al. 2022; Solomon et\\u00a0al. 2004). Importantly, the loyalty between clients and hair care professionals runs extraordinarily deep, with relationships often lasting decades and extending beyond professional services. Stylists and barbers frequently become integrated into clients\\u2019 families, attending important events and sometimes even styling deceased clients\\u2019 hair for funerals. Given the deep trust and loyalty inherent in hair care relationships, participants with Afro\\u2010textured hair may be hesitant to allow unknown researchers to manipulate their hair (Palmer et\\u00a0al. 2022). Combined with the history of stigmatization, participants may be understandably nervous, distrustful, and even alienated by well\\u2010meaning but poorly informed researchers. This underscores the importance of developing culturally responsive approaches for fNIRS research and ideally, as described below, the inclusion of research team members who share similar values and positionality.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0001\", \"dev70134-bib-0016\", \"dev70134-bib-0021\", \"dev70134-bib-0029\", \"dev70134-bib-0031\", \"dev70134-bib-0032\", \"dev70134-bib-0033\", \"dev70134-bib-0042\", \"dev70134-bib-0059\", \"dev70134-bib-0063\", \"dev70134-bib-0042\", \"dev70134-bib-0001\", \"dev70134-bib-0001\"], \"section\": \"Prior Efforts Toward More Inclusive fNIRS Research\", \"text\": \"Promisingly, many scholars have recently published calls to action and specific recommendations to increase diversity, equity, and inclusion in human neuroscience, especially in EEG, which uses a cap and scalp\\u2010based sensors similar to fNIRS (Adams et\\u00a0al. 2024; Etienne et\\u00a0al. 2020; Garcini et\\u00a0al. 2022; La Scala et\\u00a0al. 2023; Margolis et\\u00a0al. 2025; Mlandu et\\u00a0al. 2024; Murray et\\u00a0al. 2021; Richardson 2021; Webb et\\u00a0al. 2022; Wu et\\u00a0al. 2024). For example, Lietsel Richardson's \\u201cA Guide to Hair Preparation for EEG Studies\\u201d includes specific subject preparation recommendations, including style suggestions (braided crown, tight low bun, pigtail braids), and scheduling considerations, such as scheduling data collection to coincide with washday for removable styles or right before a retwist session for nonremovable styles (Richardson 2021). This may be especially relevant in pediatric populations, when investigators commonly reach out to families close to a child's birthday when they age into a new condition or project (Adams et\\u00a0al. 2024). However, birthdays and other major life events often coincide with special, and sometimes expensive, hairstyles being put in place (\\u201cbirthday hair\\u201d) (Adams et\\u00a0al. 2024). These considerations underscore the importance of communication between the research team and research participants to maximize comfort and uptake.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0008\", \"dev70134-bib-0013\", \"dev70134-bib-0015\", \"dev70134-bib-0025\", \"dev70134-bib-0065\"], \"section\": \"Prior Efforts Toward More Inclusive fNIRS Research\", \"text\": \"Inclusive practices specifically for fNIRS research\\u2014and especially pediatric fNIRS\\u2014are much more limited. Although some researchers have recently discussed these and other equity\\u2010related concerns with fNIRS (e.g., Cheaye and Rosen n.d.; Doherty et\\u00a0al. 2023; Eng 2024; Khan et\\u00a0al. 2012; Y\\u00fccel et\\u00a0al. 2024), solutions have largely focused on hardware re\\u2010design rather than specific best practices in participant interaction and hair preparation. Thus, we set out to develop participant\\u2010centered best practices to improve inclusion of participants with Afro\\u2010textured hair in fNIRS research. Rather than providing a simple, universally applicable solution, our work over the past several years has highlighted the importance of personalization and respect for the individual.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-tbl-0001\"], \"section\": \"Participants\", \"text\": \"Nineteen individuals with Afro\\u2010textured hair (11 adults, eight children; eight male, 11 female) participated in the present study. Child participants (ages 3\\u20135\\u00a0years) were participating in a longitudinal study of cognitive and brain development across preschool ages, and thus, the present investigation took place as a part of their planned fNIRS session in which they completed touchscreen\\u2010based tasks of language, theory of mind, and executive functioning skills. Adult participants were friends, family, and colleagues of author A.S. who were specifically invited to participate in this investigation. All participants identified as Black/African American with Afro\\u2010textured hair of different styles. Table\\u00a01 details participants\\u2019 sex, hair type/texture, porosity, wash frequency, and day\\u2010of\\u2010visit hairstyle. Unfortunately, signal quality data for Participant C1 were lost and thus not included in analyses, but we leave their information in the table for completeness. All participants (or their parent/guardian) provided informed consent and were compensated $20/h.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-fig-0003\"], \"section\": \"fNIRS System and Montage Specifications\", \"text\": \"Data were acquired using a NIRx NIRSport2 mobile fNIRS system with 32 dual\\u2010tip LED optodes (16 sources, 16 detectors). The montage was designed to cover brain regions associated with language, executive functioning, and social cognition, which includes bilateral prefrontal, superior/middle temporal, and temporoparietal regions (Figure\\u00a03, left). Ultimately, this created 44 topographic data channels of interest between adjacent sources and detectors (22 per hemisphere). Optodes were populated into the appropriately sized EasyCap, ranging in size from 52\\u00a0cm (child participants) to 60\\u00a0cm (adult male). The initial optode population used spring\\u2010loaded tension level 1 on the forehead (anterior four optodes on each side), tension level 2 on the front/middle of the head (the next six optodes), and tension level 3 on the back of the head (posterior six optodes); this configuration was often altered (see below).\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-fig-0003\", \"dev70134-supinfo-0001\", \"dev70134-supinfo-0001\"], \"section\": \"fNIRS System and Montage Specifications\", \"text\": \"For 15 participants (including all seven children and eight adults), one detector (F4, in right superior frontal) was sacrificed and replaced with eight short\\u2010distance detectors that were distributed evenly throughout the montage, whose purpose is to collect nonneural physiological signals for regressing out of the data from the primary channels. This left 42 long\\u2010distance data channels, plus eight short\\u2010distance data channels (Figure\\u00a03, right). Because there was less variation in the short\\u2010distance channels, we primarily examined the improvement of the long\\u2010distance data channels (see Supporting Information S2 and Figure S1 for short\\u2010distance channels).\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-fig-0004\", \"dev70134-bib-0035\"], \"section\": \"Signal Optimization\", \"text\": \"During the calibration process, the NIRSport2 device increases the brightness of each source in a stepwise manner (from 0% to 100%) until the optimal signal amplitude for each channel is obtained. The Aurora software then provides a stop\\u2010light\\u2013colored rating system for each data channel to signal researchers which channels to improve before starting the experiment (Figure\\u00a04). Red indicates critical or poor signal quality; yellow indicates acceptable (but potentially improvable) signal quality; and green indicates excellent signal quality that needs no further improvement. The colored values are based on the raw voltage measured at each detector (red <0.5\\u00a0mV; yellow\\u00a0=\\u00a0between 0.5 and 3\\u00a0mV; green >3\\u00a0mV). While these quantitative values are provided in real time during signal optimization, they change rapidly based on the physiological signal, while the coarser color categories are more stable and are typically the basis of researchers\\u2019 decisions to proceed. As stated in the Aurora user guide, \\u201cit is OK to proceed to recording with some yellow channels, but not advisable to proceed with red channels, as these will likely need to be excluded in later analysis for having a high coefficient of variation (C.V.), or noise\\u201d (NIRx Medical Technologies 2023, 25). Thus, we use the color categories as the primary outcome of interest.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-fig-0005\", \"dev70134-fig-0006\"], \"section\": \"Procedures\", \"text\": \"Below, we describe our procedures in three sections: activities before the visit, during the visit, and after the visit. We also provide two flowcharts to aid in understanding these procedures: Figure\\u00a05 describes the recruitment and scheduling process, while Figure\\u00a06 describes the explicit hair preparation options during signal optimization.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-supinfo-0001\"], \"section\": \"Before Visit: Hair Equity Screener\", \"text\": \"Based on the reviewed literature, author A.S.\\u2019s lived experience, and our team's expertise in working with children and families, we developed an intake form that we refer to as our \\u201chair equity screener\\u201d to send to participants (or pediatric participants\\u2019 caregivers) who would be engaging in fNIRS studies that require whole\\u2010head capping (i.e., this may not be necessary if studies only use prefrontal montages that only touch the forehead). Note that the present version of this screener is designed for a study of preschool children aged 3\\u20135\\u00a0years (the full text of the screener is available in Supporting Information S3). It begins with a personal introduction to the Hair Equity Specialist to help establish trust. Alongside a photo of herself, she describes her Black identity and her experience styling her own natural hair since she was 13 years old. She continues, \\u201cI hope that seeing someone who looks like you will make you more comfortable with your child being in our study. Below are more specific questions on how you take care of your child's hair.\\u201d Anecdotal participant feedback expressed gratitude for the humanizing personal details about the research team and indicated that these details have been incredibly helpful in establishing trust, providing the foundation for a relationship between participants and researchers, and offering families background information that can help ease pediatric participant anxiety through preparation before the visits.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0007\", \"dev70134-bib-0057\"], \"section\": \"Before Visit: Hair Equity Screener\", \"text\": \"Next, we asked about specific products that the caregiver uses in the child's hair, including shampoo, conditioner, and any styling products (e.g., gel, cream), as well as whether or not they ever use heat (e.g., blow\\u2010dryer). We collected information about curl type and texture and hair porosity by linking participants to an online quiz and descriptions to help them determine their types (Carol's Daughter 2023; Viola n.d.). Finally, we provided an open\\u2010field response for anything else the caregiver may want to share about their child's hair, routine, personality, or other factors. In total, this intake form establishes trust between the participant and the research team while also allowing the researchers to adequately prepare for the participant.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-tbl-0001\", \"dev70134-supinfo-0001\"], \"section\": \"Before Visit: Scheduling the Visit\", \"text\": \"If possible, participants were scheduled to come on or just before a wash day, when their hair was planned to be natural and loose. However, five adult participants were purposely invited to come while they had a protective style in place, in order to investigate possibilities for working around the style (see Table\\u00a01). After a participant completed the intake form, we scheduled the participant's lab visit. Notably, these visits typically took longer than visits for participants who do not have Afro\\u2010textured hair. While the exact timing depends on the specific montage being used, the participant population, and their tolerance for sitting still (e.g., children, individuals with disabilities), we planned for an extra hour and compensated participants for the full time they spent in the lab. In Supporting Information S4, we provide examples of infographics we provide to participants on what to expect during their lab visit.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-supinfo-0001\"], \"section\": \"Before Visit: Materials Used\", \"text\": \"We acquired a group of hair tools and products commonly used on Afro\\u2010textured hair to assist with hair preparation, including hair pics, mini elastic hair ties, scrunchies, rat tail combs, wide tooth combs, metal or plastic alligator clips, hairdryer with attachment, hair gel, hair cream, leave\\u2010in conditioner, spray bottle, barbicide, and disinfectant jar. In addition, we also used some materials specifically for fNIRS preparation, including washable chalk markers, light\\u2010up crochet hooks, clamps, and foam mannequin heads. The use of these materials is detailed below. Further, in Supporting Information S1, we provide a \\u201cshopping list\\u201d of all materials we currently keep on hand, details of how and why we use the products, and suggestions for where to purchase these items.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-fig-0006\"], \"section\": \"During Visit: Hair Preparation Procedures\", \"text\": \"Below, we describe a \\u201cmenu\\u201d of techniques that can be employed to increase the fNIRS signal with Afro\\u2010textured hair, ranging from least to most involved. These procedures were developed for use with NIRx NIRSport 2 hardware and software; however, they should be generally applicable to any fNIRS setup. In general, we aimed to use the least invasive process that achieves a good quality signal (see Figure\\u00a06 for a flowchart). Different participants required different approaches based on their hair type, their willingness and tolerance for hair styling, and the specifics of the fNIRS montage being used. Some participants required a trial of multiple approaches iteratively to achieve the best result. However, throughout the hair preparation procedure, the participant's comfort and hair integrity were prioritized, and the researcher engaged in an active assent process with each additional step or strategy attempted.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-tbl-0001\"], \"section\": \"During Visit: Hair Preparation Procedures\", \"text\": \"Interventions were applied to the whole head for child participants, as they were participating in an existing study. However, for all adult participants, we only applied the intervention to the right side of the head and kept the left side as is, so that each participant could serve as their own control. Further, the intentional inclusion of five adult participants with existing protective styles allowed us to test our full range of potential intervention options. The exact interventions used are detailed in Table\\u00a01.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-supinfo-0001\"], \"section\": \"Option #1: Gently Move the Hair\", \"text\": \"Regardless of the hairstyle that the participant arrived wearing (natural, protective style, relaxed, or other), our first approach was always to test whether we could simply gently move the hair to allow the optodes to lie flat on the participant's scalp. The cap with pre\\u2010populated optodes was fitted over the participant's head, and signal quality was monitored in real time to determine how interventions are affecting signal quality. Optodes or channels that did not have an excellent signal were flagged, and a researcher gently slid an LED\\u2010tip crochet hook (see Supporting Information S1) into a nearby, unused slit in the cap and moved the hair out of the direct path of the optode to allow the optode tip to reach the scalp. Occasionally, researchers would also try pulling the hair through nearby slits/holes in the cap (as would be done with a highlighting cap) to help keep hair out of the way. This intervention involves minimal discomfort, as the overall process is minimally invasive and the tips of the LED crochet hooks are blunted plastic.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-fig-0007\"], \"section\": \"Option #5b: Strategically Part and Braid/Twist Small Sections of Hair\", \"text\": \"When a participant had longer hair, tying or pinning the hair sections often did not sufficiently keep the volume of hair away from the parts. In this case, we parted the hair in the same way as Option 5a and then either cornrowed or flat\\u2010twisted the hair sections as flat to the scalp as possible (Figure\\u00a07). Specifically, we started cornrows or twists near the optode\\u2010dense regions and worked away toward regions without optodes, since the cornrows/twists became fatter as more hair was added and thus could push the optodes away from the scalp. While this was the most time\\u2010intensive hair preparation option (it often took 30\\u00a0min to an hour), it was also the most durable design, which may be preferred for longer studies where the fNIRS needs to be worn for a longer duration.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-fig-0008\"], \"section\": \"Option #6: Have the Participant Strategically Part and Braid/Twist the Hair at Home\", \"text\": \"For one child participant who was extremely tender\\u2010headed and apprehensive about having a stranger work with their hair, we instructed the participant's parent to create the study\\u2010specific braiding at home or at a salon before coming for their research visit. While this limited the precision that could be achieved by the experienced researchers, it was deemed the only option to include this participant. The Hair Equity Specialist set up a video call with the child's parent to describe options, demonstrated where the parts would be on a mannequin head (Figure\\u00a08), described the measurements, and showed\\u00a0examples of prior participants (faces obscured). Ultimately, this resulted in partially usable data, as some braids were aligned with the montage, but some were slightly misaligned, such that the optode landed on top of a braid rather than a part. Researchers did not attempt to rebraid the hair.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0035a\"], \"section\": \"During Visit: Realtime Signal Evaluation\", \"text\": \"Once the hair intervention was completed, we ran another signal optimization and took a screenshot. If additional interventions were needed, we repeated the process and took a final optimization screenshot after all interventions were complete. The total length of the hair intervention ranged from 30\\u00a0min to 3\\u00a0h. We also took photographs of the participant wearing the cap and additionally took a 3D whole head image using a Structure Sensor Pro (Occipital 2022) to use for post hoc evaluation of optode locations. Lastly, we removed the cap and took a final set of photographs (front/back, sides, and aerial) to document the final hair equity intervention procedure for recordkeeping. In addition to these photographs, we also kept written records of the participants\\u2019 skin color, hair color, hair texture, and hair style.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-fig-0009\", \"dev70134-fig-0009\"], \"section\": \"Results\", \"text\": \"From preintervention to postintervention (Figure\\u00a09, top), there was a significant increase in the number of green channels (t(12)\\u00a0=\\u00a04.58, p\\u00a0<\\u00a00.001; mean increase\\u00a0=\\u00a05.85 channels) and a significant decrease in the number of red channels (t(12)\\u00a0=\\u00a0\\u22125.79, p\\u00a0=\\u00a00.001; mean decrease\\u00a0=\\u00a05.69 channels). There was no significant change in the number of yellow channels. Among participants who had intervention applied to only half of the head (Figure\\u00a09, bottom), the intervened side (vs. non\\u2010intervened side) had significantly more green channels (W\\u00a0=\\u00a00.0, p\\u00a0=\\u00a00.014; mean difference\\u00a0=\\u00a05.36 channels) and significantly fewer red channels (W\\u00a0=\\u00a055, p\\u00a0=\\u00a00.006; mean difference\\u00a0=\\u00a06.45 channels). There was no significant difference in the yellow channels between the intervened and non\\u2010intervened sides. All four significant results survived FDR correction for multiple comparisons (all adjusted p\\u00a0<\\u00a00.03).\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-tbl-0001\"], \"section\": \"Results\", \"text\": \"To facilitate a qualitative comparison across individual participants, we calculated the percent improvement in usable channels (green or yellow; see Table\\u00a01). From pre\\u2010 to postintervention, improvement in usable channels ranged from 0% to 150%, with an average of 50%. Comparing right to left, improvement in usable channels ranged from \\u22125% to 59% (note that the participant with fewer usable channels on the right/intervened side used the montage with two fewer data channels on the right). Although the sample is not large enough for formal subgroup analysis, it appears that signal improvement was greatest for participants who came with loose hair that was cornrowed, and improvement was minimal for participants with protective styles (e.g., box braids and locs). Additionally, while it appears that improvement was greater for adults versus children, children tended to have better baseline signal quality before intervention, leaving less room for improvement. Finally, visual inspection suggests that the greatest signal improvement was seen in more anterior regions, as compared to more posterior regions, where hair is often thicker.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0034\"], \"section\": \"General Recommendations for fNIRS Researchers Working With Afro\\u2010Textured Hair\", \"text\": \"Building on our experiences and findings, we offer several broader recommendations for researchers conducting fNIRS studies with participants who have Afro\\u2010textured hair. First, whenever possible, research teams should hire and meaningfully involve staff with lived experience of having, caring for, and styling Afro\\u2010textured hair. Such individuals may be students, community members, or professional stylists, and their involvement can be critical for establishing trust with participants, reducing anxiety, and minimizing unintentional harm to participants\\u2019 hair, comfort, and sense of dignity. Having individuals on research teams who hold similar identities to their participants can help ensure participants feel as comfortable as possible and may also allow participants to voice concerns with diminished fear of dismissal due to racial bias (Nicolaidis et\\u00a0al. 2010). Although non\\u2010Black researchers can learn hair\\u2010equity techniques, shared lived experience often facilitates communication and rapport in ways that technical training alone cannot. Importantly, whoever fills this role should be compensated appropriately for their time, expertise, and labor, with compensation structures that may reasonably differ depending on whether this individual is a student research assistant, a community consultant, or a professional stylist.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0025\"], \"section\": \"General Recommendations for fNIRS Researchers Working With Afro\\u2010Textured Hair\", \"text\": \"Though tool improvement has not been the focus of this investigation, we offer a few recommendations and ideas. First, the fNIRS cap design could better accommodate voluminous hair. For example, larger holes distributed throughout the cap could allow hair to be pulled through more effectively, enabling the cap to lie closer to the scalp. The optimal placement of such holes may vary depending on the study\\u2010specific optode montage, suggesting that customizable cap designs would be ideal. Additionally, given the volume and density of Afro\\u2010textured hair, it may be necessary to produce larger cap sizes overall. For instance, one commonly used brand, EasyCap (Brain Products), currently only extends to 64\\u00a0cm in circumference, which may still be too restrictive for adults with very voluminous hair. Brush\\u2010fiber\\u2013tipped optodes (e.g., Khan et\\u00a0al. 2012) may also help comb through dense or coily hair, and higher tension levels of spring\\u2010loaded grommets could further improve scalp contact. However, we caution manufacturers to prioritize participant comfort to avoid introducing pain\\u2010related inequities for Black participants.\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0028\", \"dev70134-bib-0058\", \"dev70134-bib-0005\", \"dev70134-bib-0038\", \"dev70134-bib-0045\", \"dev70134-bib-0047\", \"dev70134-bib-0065\"], \"section\": \"General Recommendations for fNIRS Researchers Working With Afro\\u2010Textured Hair\", \"text\": \"Additionally, while we have focused primarily on signal attenuation due to hair, another potential source of bias in fNIRS data with Black participants is attenuation related to higher levels of melanin in the skin. Existing fNIRS technologies generally assume a fixed light pathlength through the brain across participants; however, because melanin is highly absorbent, individuals with darker, more melanated skin may experience systematic, nonlinear reductions in signal strength. This attenuation could lead to inaccurate estimations of relative changes in oxygenation in regions of interest (Kwasa et\\u00a0al. 2023; Wassenaar and van den Brand 2005). A full discussion of melanin\\u2010related limitations and methodological solutions is beyond the scope of this paper, but readers are directed to prior and emerging work in this area (e.g., Bronkhorst et\\u00a0al. 2019; Patel et\\u00a0al. 2024; Roy et\\u00a0al. 2024; Sardar et\\u00a0al. 2001; Y\\u00fccel et\\u00a0al. 2024).\"}, {\"pmc\": \"PMC12877426\", \"pmid\": \"41645562\", \"reference_ids\": [\"dev70134-bib-0033\", \"dev70134-bib-0056\"], \"section\": \"Conclusion\", \"text\": \"In conclusion, this investigation aimed to develop and evaluate best practices for conducting fNIRS research with participants who have Afro\\u2010textured hair. Without efforts at multiple system levels to diversify the research workforce, renounce scientific racism and exclusion, and repair damaged relationships with minoritized communities, calls to simply diversify samples fall short (Murray et\\u00a0al. 2021; Varma et\\u00a0al. 2021). As such, we hope that this investigation opens conversations about research practices at several levels of analysis.\"}]"

Metadata

"{\"source-schema-version-number\": \"2.0\", \"cover-date\": \"March 2026\", \"details-of-publishers-convertor\": \"Converter:WILEY_ML3GV2_TO_JATSPMC version:6.6.8 mode:remove_FC converted:06.02.2026\"}"