Researchers take optical coherence tomography to the next level
Optical coherence tomography (OCT) has been improved such that it can now photograph biological samples with greater contrast and resolution over a larger 3D field of view than was before achievable. The new 3D microscope could be beneficial for scientific research and ultimately lead to more precise medical diagnostic imaging.
The researchers from Duke University present the novel method in Optica, the journal for high-impact research published by Optica Publishing Group. They name it 3D optical coherence refraction tomography (3D OCRT). They demonstrate that 3D OCRT generates extremely detailed pictures with characteristics that are challenging to see with conventional OCT using a variety of biological materials.
OCT does not require contrast agents or labels in order to produce high-resolution 3D pictures. The imaging technique may be used to view many different body areas, such as the skin and interior of the ears, mouth, arteries, and gastrointestinal system, even though it is frequently employed for ophthalmology applications.
According to lead author Kevin C. Zhou, "OCT is a volumetric imaging method that is widely employed in ophthalmology and other disciplines of medicine." To address the well-known shortcomings of the imaging technology, "we devised a fresh and intriguing expansion, combining unique hardware paired with a new computational 3D image reconstruction algorithm."
According to co-leader of the study team Joseph A. Izatt, "We anticipate this technology being employed in a wide variety of biomedical imaging applications, such as in vivo diagnostic imaging of the human eye or skin." The equipment we created to carry out the approach may easily be downsized into tiny probes or endoscopes to reach the digestive system and other areas of the body.
Despite the fact that OCT has been effective in both clinical and biological research applications, it is challenging to concurrently obtain high-resolution OCT pictures over a large field of view in all directions due to basic restrictions imposed by optical beam propagation. Another issue is that the high amounts of random noise, or speckle, seen in OCT pictures might hide biomedically significant information.
The researchers employed an optical system that included a parabolic mirror to overcome these constraints. This kind of mirror is frequently used in non-imaging applications, such flashlights, where it encircles the lightbulb and concentrates the light in one direction. The sample was positioned where the light bulb would be in a flashlight, and the researchers employed an optical setup in which light was directed in the opposite way.
With this design, it was feasible to photograph the sample from a variety of viewpoints and angles. To integrate the views into a single, high-quality 3D image that corrects for distortions, noise, and other flaws, they used a complex algorithm.
The study team's co-leader Sina Farsiu remarked, "The work presented in Optica continues on our prior research by overcoming substantial engineering obstacles, both in the hardware and software, to allow OCRT to function in 3D and make it more broadly applicable. "We had to create a new method based on contemporary computational techniques that have recently emerged within the machine learning community because our system generates tens to hundreds of terabytes of data," the authors write.
By utilizing the technique to photograph a variety of biological samples, including zebrafish and fruit flies—important model species for behavioral, developmental, and neurological studies—the researchers showed the method's adaptability and broad application. To highlight the potential for medical diagnostic imaging, they also photographed mouse tissue samples from the trachea and esophagus. Without changing the sample, they were able to gather 3D fields of view of up to 75° using 3D OCRT.
According to Zhou, "OCRT is intrinsically capable of computationally establishing contrast from tissue qualities that are less observable in standard OCT, in addition to decreasing noise artifacts and compensating for sample-induced distortions." We demonstrate, for instance, that it responds to directed structures like fiber-like tissue.
Utilizing recent advancements in quicker OCT system technology and breakthroughs in deep learning that can speed up or enhance data processing, the researchers are currently looking into methods to reduce the system and make it faster for live imaging
Comments
Post a Comment