Info

Photograph: Andrew Cook's archive

Synchrotron X-rays Are A Billion Times More Powerful than Standard X-rays

Synchrotron X-rays Are A Billion Times More Powerful than Standard X-rays

Prof. Andrew Cook, a heart anatomist at University College London, has been closely involved in a recent study in Radiology that has unveiled unprecedented details about the human heart. Using a new imaging technique called Hierarchical Phase-Contrast Tomography (HiP-CT), researchers were able to capture intricate three-dimensional images of the heart at a level of detail not seen before. In this interview, we present the new imaging method and how it might improve treatment options in cardiology.

Prof. Andrew Cook leads the Centre for Morphology & Structural Heart Disease at UCL’s Institute of Cardiovascular Science / Great Ormond Street Hospital, now based at the GOSH/UCL Zayed Centre for Research into Rare Disease in Children, London, UK.

He is active in research, education, training with expertise in the structural architecture of the heart during fetal development, in congenital & acquired adult disease, and is the founder of The Heart Academy.

Current research focusses on Organ-to-Cell Molecular Imaging (OCMI) using Hierarchical Phase Contrast Tomography (HiP-CT); Deep-phenotyping of congenital and structural heart disease using synchrotron imaging and HiP-CT; the use of VR and 3D immersive environments for dissemination & education; and structural anatomy for device design.

Can you describe what a synchrotron is and how it helps in imaging the heart?

A synchrotron is like a small particle accelerator, about the size of a football stadium. It speeds up electrons to nearly the speed of light using magnets. These electrons emit X-rays, which are much more powerful—about a billion times stronger—than the X-rays used in hospitals for things like broken bones. The X-rays from synchrotrons are also more precise, with waves that are perfectly aligned, similar to laser light.

When these X-rays pass through soft tissue, such as a heart, they refract, or bend, just like light would when it goes through a transparent medium like water or glass. Refraction causes a shift in the X-rays that depends on the tissue's density. Special high-resolution cameras can detect this change, and after processing the data, we can create detailed 3D images of the heart, from the whole organ down to the cellular level. We can even distinguish previously ‘invisible structures’, like the heart's electrical wiring. In our research, we have used two techniques: X-ray Phase Contrast Imaging for small samples and Hierarchical Phase-Contrast Tomography for larger—adult-sized—hearts.

What challenges did you face when using the synchrotron?

As you might expect, the main challenges are technical and a full team is required for this sort of imaging, ranging from physicists, to specialist synchrotron beamline scientists, to engineers, and of course, specialists in heart structure. All need to work in unison to obtain the best heart images. When we first started out, a lot of these processes were manual. It took approximately half a day to image two-centimetre volume of tissue, and many additional hours to reconstruct. That was in 2017. Some six or seven years later, with optimisation of serial scans, reconstruction, and automated stitching, we can now do the same procedure in about twenty or thirty minutes.

There are then logistical challenges in getting samples safely to and from facilities based in different countries and building a team for imaging. When beamtime is allocated, the team will work 24/7, sometime for five days in a row, in order to maximise use of their time in the facility.

How big are these data?

This is also a major challenge for us to be able to handle these extremely large datasets. As an example, our smallest hearts are twenty gigabytes once reconstructed, which increases to between half and six terabytes for adult hearts scanned at twenty to eight micrometres voxel resolution.

The HiP-CT myomapping can recognize the heart down to 20 micrometres – half the width of a human hair. Why it is beneficial to see the heart in such a detail?

If we are to map the organisation of the myocardium in detail then we need high resolution images. Heart muscle cells in adults are fifteen to twenty microns in width, so we are close to single cell resolution at a voxel size of twenty micrometres. However, we need a minimum a three voxels to map orientation so myomapping results are at sixty micrometres, or three to four myocyte widths, at a minimum. We can also myomap at two microns in smaller ‘regions of interest’. And for small one-to-two centimetre biopsies, we can go even further down to 0.3 micrometres, which is equivalent to the size of a bacteria and means we can map capillaries, the smallest vessels in the heart, which are one to two microns in diameter.

Grenoble, France: European Synchrotron Radiation Facility (Shutterstock)
Grenoble, France: European Synchrotron Radiation Facility (Shutterstock)
The first two hearts imaged were from bodies donated for medical research. Can the method by applied to living patients? What clinical usage can it bring? 

Yes, the technique is ‘research only’ and is currently performed on explanted hearts or heart tissue. We very much hope, in the future, that it will be possible to transfer this to the clinic, but it will require major advances on the technical and facility side. It is already possible to produce X-rays in a similar way using lab-based machines, but the scan volume and imaging time is too long for clinical use at present. This aside, we believe our research will pave the way in terms of methods for analysing and interpreting the images of the heart for the time when that advance happens.

How?

Machine learning or AI will play a large role in this. It is also possible to use our current work to improve clinical imaging by a process known as ‘super-resolution’. If you know what an image should look like at high resolution you can interpolate lower resolution images to match and thereby obtain higher resolution images by imaging at lower resolution in faster scans.

In your recent report stents are specifically mentioned, how can this technology enhance treatment for patients with stents?

Our interest in imaging stents and other devices is two-fold. Firstly, it shows that both hard materials—the metal of the stents—and softer heart tissue can be imaged at the same time and reconstructed. This is not as easy as it sounds given that hard-tissue can produce artefacts on phase-contrast imaging. We can then use these images to perform computational models of the interaction between the stent and the surrounding heart tissue—perhaps even dynamically, as a stent is being deployed. In turn this will help scientists calculate the stresses placed on the stent, and also the heart tissue, and thereby optimise design and minimise failure.

How do you expect the HiP-CT myomapping will develop in the near future?

We believe HiP-CT has the potential to be a game-changer for understanding the heart in health and disease, not only for myocardial structure—via myomapping—but also for interpreting the conduction system and treatment of arrythmias and currently under-explored aspects of the microvasculature of the heart, including the cardiac lymphatics.

We are currently investigating only a fraction of the forty most common congenital heart diseases, as well as the normal heart, but are already finding novel information. We need large and consistent amounts of funding to do more. In the end the aim will be for the scientific community to use this information to produce simulations, both fluid-structure interactions and electrophysiologic, and to help improve existing computational models. It is by sharing information across disciplines and employing AI approaches that we will be able to use HiP-CT to address many challenges in heart disease.

Created: 21. 10. 2024 / Modified: 21. 10. 2024 / Mgr. Petr Andreas, Ph.D.