Easing utility in multimodal functional near infrared spectroscopy (fNIRS) and electroencephalography (EEG) in diverse participants
OA Version
Citation
Abstract
To investigate Neuroscience in the Everyday World (NEW), multimodal imaging with fNIRS and EEG has become increasingly popular as it allows for the simultaneous collection of electrical and hemodynamic data from a single participant. However, there are several key disadvantages of this type of multimodal imaging. The biggest challenge often reported is the reduction of scalp space, as experimenters now have two potentially whole head systems and only one participant’s scalp available for use. This causes experimenters to (potentially) compromise in their regions of interest to accommodate having two systems. The challenge of reduced scalp space is exacerbated by the second biggest challenge reported: potential crosstalk from the fNIRS sources onto the EEG signal. This causes experimenters to consider (and potentially increase) the distance between optodes and electrodes, further reducing scalp space. A more recent challenge is the discovery of Neuroracism within the wearable neuroscience community and instrumentation. This is due to reports that skin and hair types, found most on participants who identify as Black or African American, are associated with reduced signal quality in both fNIRS and EEG studies, leading to the exclusion of these participants data and reduced recruitment of these participants over time. Because of this, results using fNIRS and EEG studies largely do not consider the diverse population of the world and are skewed toward both Eurocentric and male participants. Given these challenges, there is a need to develop instrumentation and alternative hair maneuvering methods to both optimize scalp space and increase the inclusion of minority participants, resulting in more efficient noninvasive multimodal imaging that will represent a diverse population.In this dissertation, I will present our investigation into the effects of skin and hair characteristics on fNIRS signals and introduce new hardware developments to address practical limitations with multimodal fNIRS-EEG. Afterwards, I will highlight the utility of these procedural and hardware changes through the completion of a hearing-based study that incorporates the improvements into the study procedures.
We show that although skin and hair characteristics influence fNIRS data, there are additional accommodations that can be incorporated into the typical fNIRS data collection process that addresses these effects. We also show that the two leading practical limitations for multimodal fNIRS-EEG (reduced scalp space and potential crosstalk) can be alleviated through an alternative hardware development. Lastly, we combine the previous two results into one whole head multimodal fNIRS-EEG hearing based study to determine potential features for algorithms in hearing assistive devices. The findings in this dissertation eases the utility of multimodal fNIRS-EEG, and can both help usher the wearable neuroscience community into recruiting freely (i.e., with no restrictions or concerns related to racial demographics and phenotypes) and help improve the design of future multimodal fNIRS-EEG studies and instrumentation.
Description
2025
License
Attribution 4.0 International