Enhancement of Fourier domain optical coherence tomorgraphy images using discrete Fourier transform method

Abstract

In this paper, we address the problem of spectral data sampling in Fourier domain optical coherence tomography (FD-OCT). The interferometric information in a Fourier Domain OCT system is retrieved from spectral measurements made using a linear array spectrometer. In such spectrometers, spectral data are available as an array of points equally spaced in the wavelength domain. To obtain the spatial profile, the spectral data have to be converted to the frequency domain before applying the Fourier transform. The inverse relationship between these domains causes an unequal spacing of data points after the spectral data is converted to the frequency domain, resulting in the degradation of the FD-OCT images. The current practice typically utilizes zero-padding and spline interpolation to circumvent this problem. While these algorithms do improve the FD-OCT images, our investigations showed that more can be done to enhance the images. Toward this end, we propose a signal processing algorithm based on non-uniform discrete Fourier transform (NUDFT). The results of our algorithm are compared against the current algorithms on both simulated and experimental results.