Abstract
.Significance: By incorporating multiple overlapping functional near-infrared spectroscopy (fNIRS) measurements, high-density diffuse optical tomography (HD-DOT) images human brain function with fidelity comparable to functional magnetic resonance imaging (fMRI). Previous work has shown that frequency domain high-density diffuse optical tomography (FD-HD-DOT) may further improve image quality over more traditional continuous wave (CW) HD-DOT.Aim: The effects of modulation frequency on image quality as obtainable with FD-HD-DOT is investigated through simulations with a realistic noise model of functional activations in human head models, arising from 11 source modulation frequencies between CW and 1000 MHz.Approach: Simulations were performed using five representative head models with an HD regular grid of 158 light sources and 166 detectors and an empirically derived noise model. Functional reconstructions were quantitatively assessed with multiple image quality metrics including the localization error (LE), success rate, full width at half maximum, and full volume at half maximum (FVHM). All metrics were evaluated against CW-based models.Results: Compared to CW, localization accuracy is improved by >40% throughout brain depths of 13 to 25 mm below the surface with 300 to 500 MHz modulation frequencies. Additionally, the reliable field of view in brain tissue is enlarged by 35% to 48% within an optimal frequency of 300 MHz after considering realistic noise, depending on the dynamic range of the system.Conclusions: These results point to the tremendous opportunities in further development of high bandwidth FD-HD-DOT system hardware for applications in human brain mapping.
Highlights
Functional near-infrared spectroscopy is widely used in functional brain mapping[1] and neuro-monitoring at the bedside[2,3] due to its portability, low expense, and noninvasiveness compared to traditional neuroimaging modalities such as functional magnetic resonance imaging
Volumetric views of point spread functions (PSFs) reveal two primary characteristics: (1) while the localization error (LE) of the PSF increases with depth, the LE decreases with increasing modulation frequency (Table 3); (2) while the size of the PSF increases in depth, again as expected, the size of the PSF is only modestly affected by the modulation frequency (Table 3)
We report a systematic evaluation of image quality as a function of modulation frequency in FDHD-diffuse optical tomography (DOT) systems
Summary
Functional near-infrared spectroscopy (fNIRS) is widely used in functional brain mapping[1] and neuro-monitoring at the bedside[2,3] due to its portability, low expense, and noninvasiveness compared to traditional neuroimaging modalities such as functional magnetic resonance imaging (fMRI). The fNIRS technique is based on the differential absorption and scattering of NIR light Neurophotonics. By acquiring multiple spatially overlapping fNIRS source—detector pairs, one can obtain image reconstructions of three-dimensional (3D) spatial brain activation maps, a technique known as diffuse optical tomography (DOT).[5,6,7] early studies using sparse fNIRS systems demonstrated recovery of brain function with lower resolution and localization accuracy as compared to fMRI,[8,9,10] recent advancements with high-density diffuse optical tomography (HD-DOT), which uses overlapping multi-distance source-detector pairs, have demonstrated significant improvements in the resolution and localization error (LE) of image reconstructions.[11] More recent studies have shown potential further improvements in HD-DOT image quality and brain sensitivity when using high-bandwidth sinusoidal source modulation,[12,13] a technique often referred to as frequency domain. The effects of modulation frequency on image quality with HD-
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
