Table 6.

Summary of the challenges associated with using fNIRS in naturalistic settings and recommended solutions.

Challenge		Solution
Body movements	Motion artifacts	Correct through: - Wavelet-based filtering - tPCA
	Optical decoupling	- Properly secure the fNIRS probes to the head
Sunlight/Detector saturation		- Protecting caps - Device with ambient light detector
Signals’ quality deterioration/Channels inclusion criteria		- Visual inspection of signals - Exclude channels without heart rate oscillations - Exclude channels with CV>15% - Exclude non-measuring channels (e.g. flat lines)
Systemic changes		- Include longer rest periods (e.g., 2 min) - Band-pass filtering (NOTE: this removes some of the physiological noises, e.g. heart rate and respiration, but it is not effective in removing task-evoked systemic changes) - Measure additional physiological signals - Monitor participants’ movements (accelerometer or GPS) - Report results of ΔHbO2 and ΔHbR
Statistical inference/Unstructured protocols		- Apply AIDE

Note. Abbreviations: tPCA = targeted principal component analysis; SNR = Signal-to-noise ration; CV = coefficient of variation.