September 25, 2020

High contrast 3D histology demonstrated with sandpaper X-ray technique

Researchers at the University of Southampton have visualised a healthy mouse kidney with high spatial resolution and high contrast using a promising, emerging X-ray imaging technique.

The new study, which was led by scientists in Physics and Astronomy, highlights X-ray speckle-based imaging’s (SBI) potential as an easily implemented and non-invasive approach for next generation biomedical scanning using inexpensive optical elements.

SBI imaging is capable of delivering qualitative information, such as tissue micro-structure, and quantitative information, such as tissue density, by merely adding a piece of sandpaper to a conventional X-ray imaging setup.

Scientists have published their findings this month in The Optical Society’s flagship journal Optica.

Studying the microscopic structure of tissues is essential for a better understanding of organ functions and can, moreover, be key to identifying the cause, stage and progress of treatment of an illness. Conventional histology, the current gold standard for soft-tissue visualisation, requires several days of preparation and destroys samples so they cannot be used for further analysis. The process can also only produce 2D images that make it hard to accurately measure or analyse sample structures and their connectivity in 3D.

X-ray imaging can penetrate deeply into samples and visualise their inner structure without having to physically slice them, while also reconstructing a 3D volume by taking images at different rotation angles.

In X-ray phase-contrast imaging, scientists measure the refraction of the X-rays when passing through the sample. Compared to conventional X-ray imaging based on X-ray absorption, this approach is much more sensitive to small changes in density that are present in biological soft tissue.

This new research, led by Dr Marie-Christine Zdora, Dr Irene Zanette, and Professor Pierre Thibault, now of the University of Trieste, advances the exploration of SBI imaging beyond initial proof-of-principle demonstrations to an important real-world application.

The technology’s potential for virtual histology, where a sample is virtually sliced on a computer along a desired plane, is exhibited using a whole unstained, hydrated kidney of a healthy mouse imaged at a synchrotron, a large-scale facility that produces extremely bright X-rays.

Dr Zdora, Research Fellow and lead author of the study, says: “It seems hard to believe that simply adding a piece of sandpaper to a standard X-ray imaging setup can lead to such beautiful results and our medical collaborators were amazed by the image quality of our technique.

“Achieving such detail and contrast in hydrated biological soft tissue without the use of any staining agents is extremely challenging. The high image quality combined with the simple, cost-effective experimental setup make our method very promising for future applications in histopathology and biomedical research, but also in other fields such as materials science.

“At Southampton, we are currently working on implementing our technique at a laboratory X-ray source, which will make it more accessible to a wider range of users.”

The latest project builds upon six years of SBI research by the authors, four of them at Southampton, and was carried out under Dr Zanette’s Royal Society University Research Fellowship and Enhancement Award.

X-ray imaging took place at the ESRF (European Synchrotron Radiation Facility) in France, with conventional histology carried out on the same sample for comparison at the University Hospital Southampton’s Biomedical Imaging Unit. Samples were prepared at the University of Zurich.

The project’s funding sources included the Royal Society, the European Research Council and the Swiss National Science Foundation.

The team are next planning pre-clinical studies on medical conditions that will image healthy and diseased tissue before comparing the results. SBI will allow for measuring a large number of samples at high throughput with minimal sample preparation. The extracted 3D data volumes can then be analysed with automated algorithms to extract the desired information on the tissue. This will enable large-scale studies on a wide range of diseases in a much shorter timeframe and with less manual input than conventional histology.

The scientists are also extending the method to other fields such as materials science, and palaeontology.