Vis-OCT Explorer: an open-source software for visible-light optical coherence tomography data processing

Visible-light OCT provides enhanced retinal imaging resolution, yet achieving clinically dependable, high-quality images requires specialized post-processing methods. We developed Vis-OCT Explorer, an open-source, modular Python software package for vis-OCT image reconstruction and post-processing. In addition to standard spectral-domain OCT processing steps (k-space resampling, dispersion compensation, and Fourier transformation), it incorporates dual-spectrometer balanced detection, short-time Fourier transform (STFT)-based dispersion compensation coefficient optimization, and GPU-accelerated processing. We evaluated reconstruction performance using an intensity-based tissue contrast index and demonstrated that balanced detection significantly improved image quality. To further validate reconstruction reliability, we assessed retinal thickness measurement repeatability across five operators using intraclass correlation coefficient (ICC) analysis. Retinal nerve fiber layer (RNFL) and ganglion cell-inner plexiform layer (GCIPL) measurements demonstrated good repeatability across all operators (ICC > 0.8). Specifically, 87.5% of RNFL and GCIPL measurements from four macula-centered images exceeded this threshold. All RNFL measurements from four optic nerve head (ONH)-centered images also exceeded an ICC of 0.8, indicating strong cross-user repeatability. These results demonstrate that Vis-OCT Explorer delivers high-quality reconstruction that enables reliable quantitative measurements, supporting future multicenter clinical studies and vis-OCT data standardization.