AZoOptics caught up with George Gordon, Assistant Professor in the University of Nottingham's Faculty of Engineering, to find out more about optical fibers in microscale imaging.
How can optical fibers be used in microscale imaging and how has this technology developed?
Imaging through optical fiber bundles (also known as coherent fiber bundles or multicore fibers) has been used for many years in medical endoscopy and industrial inspection. These bundles are made from up to 100,000 light-guiding cores fused in a single glass fiber, which is typically 0.5mm-1mm in diameter.
Techniques have been recently developed that enable imaging through another type of optical fiber: multimode fibers. Multimode fibers are are cylindrical pieces of glass that can be as thin as a hair (0.1mm). However, their structure means they do not inherently preserve images. Light is scrambled as it propagates through the fibers, analogous to how frosted glass scrambles light - obscuring items behind it.
Over the past decade, researchers have shown that scrambling does not destroy information – it simply transforms it in a reversible way. This transformation can be measured by sending different known images into one end of the fiber and measuring the corresponding distorted images that come out of the other. This data is formed into a large matrix that can then be used to recover the original input from any arbitrary distorted output, thus enabling imaging through these ultrathin fibers.
Knowledge of this transformation matrix allows complete control of light exiting the fiber, opening up the possibility of performing advanced imaging modalities, e.g., quantitative phase imaging, fluorescence imaging, and polarimetry at the tip of the fiber.
What imaging applications do optical fibers have?
As they are very thin and flexible, optical fibers can enable imaging in previously inaccessible environments. For example, they can fit inside needles and so could be used to perform optical imaging deep inside the body (e.g., in the brain) in a minimally invasive way. This would make them suitable for clinical imaging.
In industrial inspection, they could be used in confined spaces (e.g., inside engines) to monitor structural damage or inside chemical reaction cells to monitor concentration gradients. Further, the fibers can be made from a material that is temperature resistant (sapphire fibers can operate at up to 2000°) and chemically inert (silica glass) for use in harsh environments.
You have recently patented new methods for the correction of dynamic optical distortions in fiber imaging. Could you please tell us about this?
Patented multilayer reflector stack concept that enables dynamic characterization of fiber distortion under realistic operation conditions where bending is constantly changing.
One of the greatest challenges in the field of fiber imaging has been that to characterize (and hence compensate) for distortions, you need to send light in at one end of the fiber and measure the resultant field at the other. This is very difficult when one end is in a patient or an engine because you need bulky optics for generating or detecting this light; which would compromise the ultrathin form factor.
Distortions cannot be characterized in advance because they constantly change as the fiber bends. My patent, funded by Cancer Research UK, presents a new method of overcoming this limitation by placing a specially designed stack of partial reflectors on the end of the fiber and using the reflected light to infer the fiber distortion.
The stack is designed so that modulating the interrogation wavelength alters the reflectance profile, providing sufficient diversity of information. It also uses nanostructured optical elements called metasurfaces to scramble light in a controlled way before reflecting it. This controlled scrambling can be measured in advance and is used to characterize dynamic distortion of the fiber during use accurately.
How could your patent be used and applied in real-world scenarios?
Our innovation could enable the development of ultrathin fiber imaging systems that support advanced imaging modalities to be used in applications where access is limited, but the environment is continuously changing. This covers a wide range of realistic use-cases: for example live imaging inside the body for medical diagnosis, imaging inside running engines, or imaging in moving fluids.
The next challenge is engineering the data acquisition and reconstruction algorithms towards being real-time to ensure this technology can be robustly applied in dynamic environments. To achieve this, we hope to partner with funding organizations and commercial entities.
What modalities of imaging can this technology perform?
So far, we have demonstrated quantitative phase imaging and polarimetric imaging through fibers. These modalities are useful for imaging transparent samples, mapping strain, profiling surfaces, and quantifying scattering.
We have recently published work showing that the quantitative phase and polarimetric imaging offers improved contrast for imaging early tumors by detecting increased light scattering caused by the disordered tissue microstructure.
Using quantitative phase and polarisation imaging through fibers to improve contrast for early cancer detection.
Top left: fluorescence microscopy image showing the location of a small tumor. Top right: amplitude imaging, the conventional method of imaging through fibers, provides poor contrast. Bottom left: quantitative phase imaging enables visualization of surface scattering of light caused by the disordered tissue microstructure of early tumors. Bottom right: polarimetric imaging also gives improved contrast.
Other groups have demonstrated confocal imaging, multiphoton imaging, fluorescence imaging, Raman imaging, and numerous other modalities through optical fibers. Our ultimate goal is to develop our novel fibre characterisation technique into a unified versatile platform that enables multiplexing of several imaging modalities in a single fibre.
How do you deal with fiber bending during use?
The need to compensate for dynamic bending-induced distortion during use has been one of the significant hurdles to the widespread application of this technology. Several approaches have been proposed, including modeling the fiber with high accuracy and using a ‘guidestar’ point source on the tip of the fiber.
Our approach does not require modeling of the fiber and does not require a light source at the tip – it simply uses reflections off of a special wavelength-selective stack structure, integrated onto the tip of the fiber. The stack diameter can match that of the fiber (0.1mm), and it can be less than 0.5mm thick, offering a compact form factor.
Demonstration of correction of bending-induced distortion.
Left: the input optical field to the fiber-containing amplitude and phase information. Middle: the raw optical field exiting the fiber with both amplitude and phase information scrambled. Right: using our distortion characterization approach, we can fully recover the input field in terms of both amplitude and phase.
By transmitting many different spatial light patterns across three wavelengths into one end of the fiber and recording the resultant light that travels down the fiber, reflects off the stack, travels back up the fiber and exits at the same end, we can characterize instantaneous distortion. The major remaining challenge is to build a full experimental demonstration that achieves distortion-correction at real-time speeds.
How will this technology help to influence future developments and research?
This technology will enable the field of fiber imaging to expand into new applications, particularly those involving imaging in very challenging environments with limited space, significant movement, or harsh chemicals. Additionally, it will allow cutting-edge imaging modalities to be trialed in these new environments.
This could then enable studies of cellular interactions deep within the body for monitoring the efficacy of drugs. It could also allow visualization of chemical reactions inside industrial processes for improved yield. It could even enable live examination inside working engines to increase efficiency and identify failures before they happen.
This technology, therefore, has great potential to be used as a versatile tool for future research and development.
About George Gordon
George Gordon is an Assistant Professor in the Department of Electrical and Electronic Engineering at the University of Nottingham. His research is on optics, photonics, and medical imaging, specifically developing new types of optical endoscopes that improve the detection of diseases such as cancer.
Prior to this post, he was a post-doctoral researcher and research fellow at the University of Cambridge, where he constructed a prototype holographic optical fiber endoscope that images optical intensity, phase, and polarisation. He is also interested in low-cost imaging to develop new capsule endoscopes for early cancer detection in the esophagus and plasmonic metasurfaces for low-cost multispectral imaging.
Before moving into medical imaging, George completed his Ph.D. in Telecommunications Engineering at the University of Cambridge, having done his undergraduate degree in Electrical and Electronic Engineering at the University of Auckland in New Zealand.
More info on George’s research and publications can be found at www.georgesdgordon.com.
- George S. D. Gordon, James Joseph, Travis Sawyer, Alexander J. Macfaden, Calum Williams, Timothy D. Wilkinson, and Sarah E. Bohndiek, "Full-field quantitative phase and polarisation-resolved imaging through an optical fibre bundle," Opt. Express 27, 23929-23947 (2019), Available at: https://www.osapublishing.org/oe/abstract.cfm?uri=oe-27-17-23929 [Accessed: 15 October 2019]
- George S. D. Gordon, Milana Gataric, Alberto Gil C. P. Ramos, Ralf Mouthaan, Calum Williams, Jonghee Yoon, Timothy D. Wilkinson, and Sarah E. Bohndiek, "Characterising optical fibre transmission matrices using metasurface reflector stacks for lensless imaging without distal access", arXiv:1904.02644 [physics.optics] (2019), Available at: https://arxiv.org/abs/1904.02644 [Accessed: 15 October 2019]
- George S. D. Gordon, James Joseph, Maria P. Alcolea, Travis Sawyer, Alexander J. Macfaden, Calum Williams, Catherine R. M. Fitzpatrick, Philip H. Jones, Massimiliano di Pietro, Rebecca C. Fitzgerald, Timothy D. Wilkinson, and Sarah E. Bohndiek, "Quantitative phase and polarisation endoscopy applied to detection of early oesophageal tumourigenesis", arXiv:1811.03977 [physics.med-ph] (2018), Available at: https://arxiv.org/abs/1811.03977 [Accessed: 15 October 2019]
Disclaimer: The views expressed here are those of the interviewee and do not necessarily represent the views of AZoM.com Limited (T/A) AZoNetwork, the owner and operator of this website. This disclaimer forms part of the Terms and Conditions of use of this website.