Presentation
9 March 2023 Improving the estimation accuracy of the scatterer density estimation by accounting for the spatial property of the noise in optical coherence tomography
Author Affiliations +
Abstract
A new deep-learning-based scatterer density estimator (SDE) is demonstrated. The SDE is trained by pairs of numerically simulated OCT images and its background parameters including the scatterer density, resolutions, and signal-to-noise ratio. For this simulation, we introduced a new noise model that accurately accounts for the spatial properties of three noise types: shot, relative-intensity, and detector noise. This SDE was experimentally validated by phantom and in-vitro tumor spheroid measurements. Significantly improved accuracy was found in comparison to our old SDE being trained with a naïve noise model that does not account for the spatial noise property.
Conference Presentation
© (2023) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Thitiya Seesan, Pradipta Mukherjee, Ibrahim Abd El-Sadek, Yiheng Lim, Shuichi Makita, Prathan Buranasiri, and Yoshiaki Yasuno "Improving the estimation accuracy of the scatterer density estimation by accounting for the spatial property of the noise in optical coherence tomography", Proc. SPIE PC12367, Optical Coherence Tomography and Coherence Domain Optical Methods in Biomedicine XXVII, PC123670Z (9 March 2023); https://doi.org/10.1117/12.2648798
Advertisement
Advertisement
KEYWORDS
Optical coherence tomography

Signal to noise ratio

Computer simulations

Data modeling

Neural networks

Scattering

Sensors

Back to Top