Fig. 2: Correction of sample motion enables field-based diffuse optics.
From: Time-of-flight resolved light field fluctuations reveal deep human tissue physiology
a Geometry for human forearm measurement, where αSD (source–detector angle) is ~10°. For a 2 ms segment at the peak time-of-flight (solid black vertical line in b), the autocorrelation (c) rotates with lag time τd, implying a linear phase shift due to sample motion. d The rate of phase change, i.e. the Doppler frequency shift, can be calculated from the power spectrum (Fourier transform of the autocorrelation \({{G}}_{1,{{w}}}^{{\mathrm{iNIRS}}}\)) of the segment. e The Doppler frequency shift time course can also be expressed as a Doppler velocity ∆V assuming αSD = 0°. f Motion correction using the estimated phase (green) virtually eliminates phase drift at the peak TOF, as seen by comparing black and red curves. As shown in Supplementary Fig. 2, phase correction also restores the validity of the modified Siegert relationship. g–j Simultaneous comparison of bulk phase shifts due to axial motion in iNIRS and optical coherence tomography (OCT). g The two systems are combined with a dichroic beam splitter such that ~1320 nm OCT light is transmitted, while 855 nm iNIRS light is reflected. h OCT cross-sectional image from the combined system; Doppler velocity to estimate bulk motion is calculated from the first 50 µm from the skin surface as shown by the green arrow. The Doppler velocity parametric plot (i) and time courses (j) show excellent agreement between modalities (R2 = 0.95).
