In sync: phase-contrast imaging

Observing the apparent bend of a straw in a glass of water might take most of us back to science class in primary school, but for Professor Alessandro Olivo, this law of physics has been the basis of his breakthrough research that has helped industries as varied as security, cultural heritage and, most importantly, medical imaging.

The professor of applied physics at University College London’s (UCL) Department of Medical Physics and Biomedical Engineering says what he does is simple: “It’s just refraction.”

Olivo and his 14-member team have developed phase-contrast X-ray imaging (PCXI) technology to produce high-quality images at significantly lower doses of radiation. He has essentially taken the powerful imaging capabilities previously confined to synchrotrons – highly specialised light generators or X-ray generators – and reproduced them in a laboratory setting.

Leaning back in his chair, Olivo gives off a relaxed energy. Having spent the past 20 years of his professional life studying X-ray imaging, he knows how to make complex subject matter straightforward.

While source and detector technology has evolved since Wilhelm Röntgen’s accidental discovery of X-rays 120 years ago, Olivo says they still relied on attenuation to produce images. A radiologist will look at how many X-rays are stopped by the tumour and how many by the soft tissue. The difference between those numbers will provide an image. However, when it comes to imaging a tumour amid soft tissue of a similar density, the contrast in the image flattens and may not be as clear as it needs to be.

Something different

“This is an intrinsic limitation of how X-rays are used,” Olivo says, “and the only way to solve it is to try to look at it in a completely different way.

“What we do is look at a separate physical mechanism and at phase changes. X-rays are waves and do a number of different things when they travel through materials. They get differently attenuated but, most importantly, they change speeds. This is difficult to detect, but it is effectively 1,000 times larger than the variation of attenuation.”

Olivo says his team asked, ‘What manifestation of a phase change would be easiest to detect outside of a synchrotron?’. The answer is refraction. A change in phase can manifest as a deviation in the X-ray direction at a specific point, which is equivalent to refraction. Remember the straw in the glass of water; “Exactly the same thing happens in X-rays; however, the angle is much smaller – it is scaled to the wavelength. So instead of looking at a few degrees, you’re looking at something like a micro radiant, which is the angle subtended by a millimetre-sized object placed a kilometre away,” Olivo adds.

The challenge was to devise a system that was able to pick up angular deviations on such small scales. “The result is transformative contrast that can be orders of magnitude above the conventional attenuation contrast, which has two consequences,” says Olivo. “You can see things you previously thought were invisible because they did not generate sufficient differential attenuation, but you can also increase the contrast of everything in the image.

“You can see the same thing a lower dose, or see more with the same dose, or any combination in between.”

Refraction needs masks. One is placed in front of the sample to shape the beam, dividing it into multiple beamlets. Importantly, every beamlet has to intercept a sensitivity edge. This can be achieved by placing a second mask in front of the detector, so that every beam hits the edge of an aperture in the second mask.

When a beamlet is deflected by refraction and it hits the edge, the fraction of the beam intercepted by the edge grows or decreases depending on the directional refraction. The edge stops a particular number of photons, depending on the angle, which converts an invisible phase effect into a visible intensity difference.

“All the detector understands is intensity,” Olivo says. “Hence, a mechanism must be put in place to transform an undetectable phase change into a detectable intensity variation.”

It’s the study of detectors that led Olivo to X-ray imaging. After graduating from Italy’s University of Trieste, where he also completed his PhD, he began his career as a high-energy physicist. He felt like “a small wheel in a gigantic mechanism. They pass you bits of data; you do some analysis and you pass them to someone else. I wanted to work in an environment where you could get the bigger picture and know what was going on across the entire project. So that took me to detector development.”

Adapt and thrive

The same detectors used in high-energy physics to spot particles in accelerators have been adapted and converted to be used as imaging devices. “That is pretty much what I’ve done,” he says. “I started developing detectors and then realised that it is a great idea to use them for imaging.”

Olivo dedicated his initial research efforts to developing single-photon counting edge-on silicon microstrip detectors, which now form the basis of the SECTRA mammography system. The timing was rather fortuitous as a synchrotron was being built in Italy.

It was then that Olivo saw the “perfect match” between the detectors he was developing and the light properties at the synchrotron. “We started to use them together, and I stumbled into these phase effects,” he says. “I had the idea they could change the way people do X-ray imaging, and I took it from there.”

Best of both worlds

Olivo’s work has been a significant breakthrough for multiple industries; however, in medical imaging, the major boon has been for diagnosing cancer. MRI scanners are excellent for these types of images; however, they are expensive and bulky, and don’t offer the same high resolution as X-rays without having long acquisition times. X-rays are cheaper and faster, but they don’t see soft tissue.

“We wanted the best of both worlds,” Olivo says. “If we were able to see soft tissue differences using X-rays then we’d have everything: compact, cost-effective and speedy X-rays combined with the soft tissue sensitivity hopefully comparable to that of MRI.”

Historically, when diagnosed with a breast tumour, a patient needed a mastectomy, which requires removing the entire breast. These days, doctors prefer lumpectomies or breast-conservation surgery where the surgeon removes only the tumour, preserving most of the breast, limiting the plastic surgeon’s subsequent interventions. The surgeon has to make sure the tumour does not reach the removed tissue’s surface margins.

“That takes time; you cannot do it there and then,” Olivo explains. “You send samples of this lump to the histopathology lab and they do the analysis, which is very accurate and precise, but the answer is received after a couple of weeks.

“At that point, if there is an involved margin, you have to call the woman back. So they have already suffered a
lot of stress, had an operation, and now have to be told they still have cancer and need another operation. This is terrible.”

Olivo’s PCXI technology aims to use its increased sensitivity to soft tissue differences to determine the extent of the tumour and whether it reaches the margins or not. “The answer is received in minutes,” Olivo says. “The surgeon might continue to remove tissue until confident that the entire tumour has been removed, and we hope this will cut down on the reoperation rate by a big amount. Only the project will tell us by how much, but I hope it’s more than a 50% reduction in the rate.

In the longer term, Olivo hopes to develop his PCXI technology into an in-vivo diagnostic tool. As well as potentially enabling earlier detection of a range of life-threatening diseases, there is scope for significant reduction in radiation exposure – a concept for which Olivo’s team has recently achieved preliminary data.

“What we’ve done is take ex-vivo breast tissue, gone to the synchrotron, optimised the X-ray energy to much higher than you’d normally use in mammography – three of four times higher – and we have shown that by using that – the phase instead of the attenuation – you can get at least the same if not better image quality with a dose reduction by a factor of about 20.”

With this type of technology, Olivo says breast screening becomes almost harmless, which means women could be checked at younger ages and more frequently. The power of XPCI extends beyond tumours to imaging bone and cartilage, and even degenerative diseases.

“The technology can be an enabler for new treatment and a new way in which we go about curing people,” Olivo says.

So when will this technology be ready for use in radiologists’ labs and hospitals? As always, academic and medical progress needs investment to take discoveries further. “It very much depends on how prepared big companies are to take a risk,” Olivo says. “At the moment, some of them almost see this as science fiction.

“Companies are very risk-averse after 2008 [financial crisis]. If you did this 20 years ago, I think it would have been a no-brainer and people would have taken it up straight away.”

Olivo’s hope is that this technology could be developed into X-ray machines for clinical settings within five years. This is particularly the case for areas where there is a lot of existing data, such as mammography.

Olivo and his team have made a major breakthrough. They have done the research and demonstrated the principles that highlight the evolution of the technology. There is work left to do, but he is confident investors will be brave enough to move this technology to the next stage of development.