Signal Processing in fNIRS: A Case for the Removal of Systemic Activity for Single Trial Data

Abstract

Researchers using functional near infrared spectroscopy (fNIRS) are increasingly aware of the problem that conventional filtering methods do not eliminate systemic noise at frequencies overlapping with the task frequency. This is a problem when signals are averaged for analysis, even more so when single trial data are used as in online neurofeedback or BCI applications where insufficiently preprocessed data means feeding back noise instead of brain activity or when looking for brain-behavior relationships on a trial-by-trial basis. For removing this task-related noise statistical approaches have been proposed. Yet as evidence is lacking on how these approaches perform on independent data, choosing one approach over another can be difficult. Here signal quality at the single trial level was considered together with statistical effects to inform this choice. Compared were conventional band-pass filtering and wavelet minimum description length detrending and the combination of both with a more elaborate, published preprocessing approach for a motor execution—motor imagery data set. Temporal consistency between Δ[HbO] and Δ[HbR] and two measures of the spatial specificity of signals that are proposed here served as measures of data quality. Both improved strongly for the combinationed preprocessing approaches. Statistical effects showed a strong tendency toward getting smaller for the combined approaches. This underlines the importance to adequately deal with noise in fNIRS recordings and demonstrates how the quality of statistical correction approaches can be estimated.

The potential of functional near-infrared spectroscopy-based neurofeedback – a systematic review and recommendations for best practice

“…On the other hand, (Kirilina et al, 2012) reported HbR to be less sensitive to artifacts. However, a recent study (Klein and Kranczioch, 2019) demonstrated that HbR is also affected by a global signal component. The contrast-to-noise ratio seems to be comparable for HbO and HbR across different tasks (Cui et al, 2011), but according to Naseer and Hong (2015) HbO signals were more discriminative for BCI applications than those of HbR signals.…”

“…Short-distance channels in combination with GLM seem to be the most efficient tool to correct for extracerebral physiological signal components (Brigadoi and Cooper, 2015;Tachtsidis and Scholkmann, 2016;von Lühmann et al, 2020). As already stated, only Fujimoto et al (2017) used this technique, which may be because most of the fNIRS systems are not equipped with the appropriate hardware (Klein and Kranczioch, 2019). If this is the case, a potential alternative is the global component removal approach as introduced by Zhang et.…”

“…(2016). This technique seems to be promising to reduce global physiological signals from the fNIRS data and can be used for online artifact control as recently pointed out by (Klein and Kranczioch, 2019) specifically with respect to single-trial data. FNIRS-neurofeedback studies have not yet applied this method but have rather used CAR or other referencing to correct for evoked systemic cerebral and extracerebral components.…”

The Potential of Functional Near-Infrared Spectroscopy-Based Neurofeedback—A Systematic Review and Recommendations for Best Practice

“…Short-distance channels in combination with GLM seem to be the most efficient tool to correct for extracerebral physiological signal components (Brigadoi and Cooper, 2015 ; Tachtsidis and Scholkmann, 2016 ; von Lühmann et al, 2020 ). As already stated, only Fujimoto et al ( 2017 ) used this technique, which may be because most of the fNIRS systems are not equipped with the appropriate hardware (Klein and Kranczioch, 2019 ). If this is the case, a potential alternative is the global component removal approach as introduced by Zhang et al ( 2016 ).…”

Differences in Net Information Flow and Dynamic Connectivity Metrics Between Physically Active and Inactive Subjects Measured by Functional Near-Infrared Spectroscopy (fNIRS) During a Fatiguing Handgrip Task

“…An alternative method would be to employ short-distance channels (<1 cm), which were not available in the commercial optode holder cap used, to remove by regression hemodynamic fluctuations that co-occur in the cortex as well as superficial scalp layers (Gagnon et al, 2012;Tachtsidis and Scholkmann, 2016). Although we used bandpass filters and PCA to remove physiological inferences and global hemodynamic fluctuations, there exists several other different computational methods including (i) independent component analysis, (ii) singular value decomposition (SVD) and Gaussian kernel smoothing, (iii) statistical correction methods, (iv) wavelet-based methods, or (v) a combination of these methods can be used for removal (Huppert et al, 2009;Jang et al, 2009;Tachtsidis and Scholkmann, 2016;Cao et al, 2018b;Duan et al, 2018;Klein and Kranczioch, 2019). Thus, a quantitative comparison using different global component removal methods is warranted in future studies.…”