The axial resolution is an important parameter in Optical Coherence Tomography (OCT). In OCT a
broadband light source is used to achieve high axial resolution imaging. However the dispersion results
in a broadening of the coherence envelope. The dispersion mismatch between reference and sample
arms then needs to be minimized to achieve optimal axial resolution for OCT. In this work we propose
a new numerical dispersion compensation method to obtain ultrahigh resolution in SDOCT, in which
wavelet transform instead of Fourier transform is used to obtain the signal in different frequency
domain. And a series of the phase signals of different interfaces of the sample can be obtained. Under
the homogeneous medium approxiamtion, the phase signal is a linear function of the wave number.
Thus based on linearization of the phase signal of different interface and the wave number, the axial
resolution can be improved.
A different real-time self-wavelength calibration method for spectral domain optical coherence tomography is presented in which interference spectra measured from two arbitrary points on the tissue surface are used for calibration. The method takes advantages of two favorable conditions of optical coherence tomography (OCT) signal. First, the signal back-scattered from the tissue surface is generally much stronger than that from positions in the tissue interior, so the spectral component of the surface interference could be extracted from the measured spectrum. Second, the tissue surface is not a plane and a phase difference exists between the light reflected from two different points on the surface. Compared with the zero-crossing automatic method, the introduced method has the advantage of removing the error due to dispersion mismatch or the common phase error. The method is tested experimentally to demonstrate the improved signal-to-noise ratio, higher axial resolution, and slower sensitivity degradation with depth when compared to the use of the zero-crossing method and applied to two-dimensional cross-sectional images of human finger skin.