STORM-Net: Simple and Timely Optode Registration Method for Functional Near-Infrared Spectroscopy (fNIRS)
AbstractSignificanceWe propose a robust video-based method for estimating the positions of fNIRS optodes on the scalp.AimCalibrating the location of optodes relative to a subject’s scalp is an important step in acquisition of reliable neuroimaging data, and is a relatively open problem when dealing with developmental populations. Existing methods pose various motion constraints, require expert annotation and are only applicable in laboratory conditions. A quick and robust framework to deal with these issues is required.ApproachUsing a variety of novel computer-vision technologies, we implement a fully-automatic appearance-based method that estimates the registration parameters from a raw video of the subject. We validate our method on 10 adult subjects and prove its usability with infants as well.ResultsWe compare our method with the golden standard 3D digitizer, and to other photogrammetry based approaches. We show it achieves state-of-the-art results. Our method is implemented as a freely available open-source toolbox at https://github.com/yoterel/STORM.ConclusionsOur method allows to calibrate the fNIRS system in a simple way, with unprecedented speed and accuracy. Fast calibration facilitates more spatially precise neuroimaging with developmental and clinical populations even in unconventional environments.