Fast and high-resolution optical coherence tomography reconstruction via a resampling-free spectral fitting framework

Speed and axial resolution are the key factors that determine the widespread application of OCT. FFT, which is the traditional algorithm used in OCT, has the disadvantages of requiring pre-processing such as resampling and being limited by the spectrum of the light source, which limits the speed and axial resolution. This paper proposed a spectral fitting-based method for fast and high-resolution reconstruction in optical coherence tomography (OCT). The proposed method establishes a model to derive a linear relationship between the interferometric spectrum and imaging depth. By constructing a spectral fitting framework of the interferometric signal, a system coefficient matrix is obtained. The depth profile can then be efficiently reconstructed through matrix multiplication between the matrix and the measured spectra. The comparison of the results obtained by this method, the autoregressive (AR) method, and the iterative adaptive approach (IAA) demonstrated that the proposed method exhibits great potential for fast and high-resolution OCT imaging and is expected to be applicable to real-time biomedical imaging.