Presentation
5 March 2021 Rotational distortion compensation with deep learning for proximal-scanning endoscopic optical coherence tomography
Guiqiu Liao, Oscar Caravaca Mora, Philippe Zanne, Benoit Rosa, Diego Dall'Alba, Paolo Fiorini, Michel de Mathelin, Florent Nageotte, Michalina Gora
Author Affiliations +
Abstract
The rotational distortion of endoscopic Optical Coherence Tomography (OCT) is caused by friction of optical fiber and motor instabilities. On-line rotational distortion compensation is essential to provide real-time feedback. We proposed a new method that integrates a Convolutional Neural Network based warping parameters prediction algorithm to correct the azimuthal position of each image line. This method solves the problem of drift in iterative processing by an overall shifting parameter predicting nets with a processing time of 145ms/frame and variation reduction of 88.9% for the data obtained in ex-vivo and in-vivo experiments.
Conference Presentation
© (2021) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Guiqiu Liao, Oscar Caravaca Mora, Philippe Zanne, Benoit Rosa, Diego Dall'Alba, Paolo Fiorini, Michel de Mathelin, Florent Nageotte, and Michalina Gora "Rotational distortion compensation with deep learning for proximal-scanning endoscopic optical coherence tomography", Proc. SPIE 11620, Endoscopic Microscopy XVI, 1162005 (5 March 2021); https://doi.org/10.1117/12.2576882
Advertisement
Advertisement
KEYWORDS
Optical coherence tomography

Distortion

Endoscopy

Robotics

Video

Detection and tracking algorithms

Evolutionary algorithms

Back to Top