Full resolution Fourier domain optical coherence tomography

Abstract

The complex conjugate ambiguity in Fourier domain optical coherence tomography (FDOCT) is a major roadblock that prevents full-detector resolution for the depth scan in single shot operation for the individual A-scan. Current techniques to eliminate this problem involve changing the experimental set-up, usually complicating the OCT system. In this work we show that the standard FDOCT spectrum data when resampled appropriately can be cast exactly in terms of type-1 discrete cosine transform (DCT). Additionally, a sparse reconstruction method in the DCT domain enables image recovery with full-detector resolution, thus effectively doubling the depth scan resolution. In a realistic simulation study we demonstrate full-detector resolution for a discrete reflective target by successfully resolving closely spaced reflective peaks that cannot be separated using the standard Fourier transform based reconstruction. Experimental results on reflective glass sheet targets further validate the methodology. The results of the proposed technique suggest that full resolution FDOCT systems may be implemented practically without additional hardware costs and system complexity.

Wavelength-encoded multi-channel spectral domain optical coherence tomography (SD-OCT) for spatiotemporally resolved imaging

“…A direct current (DC) noise was removed from the raw composite interferogram, I c (λ), using reference subtraction and high-pass filtering. This was followed by a resampling process that involved Fourier Thus far, the computational complexity of the signal processing is equivalent to that of a standard single-channel OCT system in which the DC removed intensity, I single (λ), of the spectral interferogram for N discrete reflecting layers in the sample can be expressed by equation (1) [29,30]:…”