Optimized linear wavenumber spectrometer based spectral-domain optical coherence tomography system

• DOI: 10.7498/aps.67.20172606
• Publication Date: 2020-12-21T02:48:56Z
• Authors: Wu Tong; Sun Shuai-Shuai; Wang Xu-Hui; Wang Ji-Ming; He Chong-Jun; Gu Xiao-Rong; Liu You-Wen
• Keywords: Wavenumber; Imaging spectrometer

In spectraldomain optical coherence tomography the sample is illuminated by a broadband light source, and spectrum of interference between returned from reference mirror detected grating spectrometer. Conventionally, spectrometer comprised diffraction grating, focusing lens, line-scan camera. According to equation angle approximately linearly related wavelength. Thus distribution function at camera nonlinearly dependent on wavenumber. For high-quality image reconstruction, numerical resampling spectral data wavelength-space wavenumber-space commonly required prior Fourier Transformation. The nonlinear detection interferograms in wavenumber space also degrades depth-dependent signal sensitivity conventional linear-wavelength based tomography. Recently reported linearwavenumber does not need or interpolating nonlinearwavenumber data, which greatly reduces cost computation improves imaging sensitivity. Various methods different evaluation protocols for optimizing design linear-wavenumber have been reported. Here we report an effective optimization method used high-resolution domain system. We take reciprocal fullwidthhalfmaximum simulated point spread as evaluating criterion optimize structure parameters spectrometer, including refractive index vertex dispersive prism rotation prism. optimization, F2 equilateral construct optimized with 21.8°. implement developed system unit. evaluate performances both theoretically experimentally. experimentally measured axial resolution 8.52 μm, be 91 dB -6 roll-off within depth range 1.2 mm. curve accords well theoretical curve. Utilizing general parallel computing capability GPU card, highquality images human finger skin can reconstructed real time without any process.