Dual-slope imaging of cerebral hemodynamics with frequency-domain near-infrared spectroscopy

SIGNIFICANCE: This work targets the contamination of optical signals by superficial hemodynamics, which is one of the chief hurdles in non-invasive optical measurements of the human brain. AIM: To identify optimal source–detector distances for dual-slope (DS) measurements in frequency-domain (FD) near-infrared spectroscopy (NIRS) and demonstrate preferential sensitivity of DS imaging to deeper tissue (brain) versus superficial tissue (scalp). APPROACH: Theoretical studies (in-silico) based on diffusion theory in two-layered and in homogeneous scattering media. In-vivo demonstrations of DS imaging of the human brain during visual stimulation and during systemic blood pressure oscillations. RESULTS: The mean distance (between the two source–detector distances needed for DS) is the key factor for depth sensitivity. In-vivo imaging of the human occipital lobe with FD NIRS and a mean distance of 31 mm indicated: (1) greater hemodynamic response to visual stimulation from FD phase versus intensity, and from DS versus single-distance (SD); (2) hemodynamics from FD phase and DS mainly driven by blood flow, and hemodynamics from SD intensity mainly driven by blood volume. CONCLUSIONS: DS imaging with FD NIRS may suppress confounding contributions from superficial hemodynamics without relying on data at short source–detector distances. This capability can have significant implications for non-invasive optical measurements of the human brain.