Automatic motion compensation for structured illumination endomicroscopy using a flexible fiber bundle

Abstract

Significance: Confocal laser scanning enables optical sectioning in clinical fiber bundle endomicroscopes, but lower-cost, simplified endomicroscopes use widefield incoherent illumination instead. Optical sectioning can be introduced in these simple systems using structured illumination microscopy (SIM), a multiframe digital subtraction process. However, SIM results in artifacts when the probe is in motion, making the technique difficult to use in vivo and preventing the use of mosaicking to synthesize a larger effective field of view (FOV). Aim: We report and validate an automatic motion compensation technique to overcome motion artifacts and allow generation of mosaics in SIM endomicroscopy. Approach: Motion compensation is achieved using image registration and real-time pattern orientation correction via a digital micromirror device. We quantify the similarity of moving probe reconstructions to those acquired with a stationary probe using the relative mean of the absolute differences (MAD). We further demonstrate mosaicking with a moving probe in mechanical and freehand operation. Results: Reconstructed SIM images show an improvement in the MAD from 0.85 to 0.13 for lens paper and from 0.27 to 0.12 for bovine tissue. Mosaics also show vastly reduced artifacts. Conclusion: The reduction in motion artifacts in individual SIM reconstructions leads to mosaics that more faithfully represent the morphology of tissue, giving clinicians a larger effective FOV than the probe itself can provide.

High-Resolution Image Processing of Probe-Based Confocal Laser Endomicroscopy Based on Multistage Neural Networks and Cross-Channel Attention Module

“…This proves especially beneficial for confocal imaging, as it allows the laser scanning system to be positioned outside the patient [2]. The systems rely on fiber optics, employ laser technology, and necessitate rapid and accurate scanning mechanisms [3].…”

Real-timing processing of fiber bundle endomicroscopy images in Python using PyFibreBundle

“…A non-sectioning widefield endomicroscope can easily be built using a fiber bundle [7], and with a careful choice of topical fluorescent contrast agents this offers an effective and lowcost approach to in vivo microscopy for applications such as detecting oral cancer [8]. Line-scanning [9,10] and structuredillumination endomicroscopes [11][12][13] trade off partial optical sectioning in exchange for higher frame rates. Other techniques include multi-spectral imaging [14], two photon microscopy [15], fluorescence lifetime imaging [16], holographic microscopy [17,18], light field imaging [19], and optical coherence tomography [20].…”

Real-timing processing of fiber bundle endomicroscopy images in Python using PyFibreBundle

“…A non-sectioning widefield endomicroscope can easily be built using a fiber bundle [7], and with a careful choice of topical fluorescent contrast agents this offers an effective and low cost approach to in vivo microscopy for applications such as detecting oral cancer [8]. Line-scanning [9,10] and structured-illumination endomicroscopes [11][12][13] trade off partial optical sectioning in exchange for higher frame rates. Other techniques include multi-spectral imaging [14], two photon microscopy [15], fluorescence lifetime imaging [16], holographic microscopy [17,18], light field imaging [19] and optical coherence tomography [20].…”