Abstract
Visible light optical coherence tomography (VIS-OCT) provides retinal oximetry at micro-level vessels by performing spatiospectral analysis. Typical methodology involves the short-time Fourier transform (STFT), which requires computationally intensive repetitive transforms. Here we report a depth-gated Fourier transform (DGFT) method to reduce the number of transforms (and time) for spectral extraction by windowing the depth domain. The number of transforms was decreased from 13 to 3 by DGFT, nearly 6× faster in computation time than STFT. We validated DGFT for retinal oximetry in a human eye. Oxygen saturation (sO(2)) values matched well between STFT and DGFT (percent difference of 0.63% ± 1.10%), while the DGFT extracted spectra significantly faster than the STFT (0.15 ± 0.11 s vs 0.89 ± 0.48 s). The reported method shows potential for real-time oximetry calculation in the future.