When Natalie Coutts鈥� nephew, sister and uncle died from cancer in the space of five years, doctors were puzzled.

They seemed far too young 鈥� her nephew, who died of a brain tumour, was only 6. Her sister died at just 29 from soft-tissue cancer, and her uncle at 49 from pancreatic cancer.

鈥淭he doctors told us, 鈥業t isn鈥檛 normal for so many young people in one family to pass away from cancer鈥�,鈥� says Coutts, who is now 41 and lives in Melbourne.

When some of her relatives underwent genetic testing in the early 2000s, they discovered they had Li-Fraumeni syndrome 鈥� a rare hereditary condition caused by mutations in a gene called TP53. The mutations, which occur in about 1 in 10,000 people, greatly increase cancer susceptibility. Half of carriers develop cancer by the age of 30, and the overall lifetime risk is almost 100 per cent.

Coutts鈥� father 鈥� who recently died of leukaemia 鈥� turned out to be a carrier. So did her aunt, who recently lost her life to lung cancer. Two of her siblings also tested positive 鈥� one remains cancer-free and the other died of other causes.

Test quandary

For a long time, Coutts was unsure about doing the genetic test. 鈥淚 um-ed and ah-ed for five years,鈥� she says. 鈥淲hen you don鈥檛 know, you can cling on to the possibility that you might be negative,鈥� she says. 鈥淏ut that limbo is also very unpleasant, so in the end, I decided I needed to know.鈥�

When she finally took the test in 2009, at the age of 33, she was confirmed as having a TP53 mutation. 鈥淚t was absolutely devastating,鈥� she says.

But now, Coutts鈥� prospects are looking brighter. In 2012, she joined a trial at the Garvan Institute of Medical Research in Sydney with 29 other people who have Li-Fraumeni syndrome. The aim of the study was to see if having a whole-body MRI scan each year could detect cancers before they turned deadly.

So far, the screening technique has spotted tumours in Coutts鈥� uterus and kidneys. They turned out to be benign, but Coutts is glad they were picked up and removed anyway. 鈥淚 don鈥檛 care how many times I鈥檓 under the knife because I have children and I want to be around for them,鈥� she says. 鈥淚 want every lump and bump investigated because there鈥檚 such a high risk of cancer.鈥�

Medical treadmill

In addition to whole-body MRI, Coutts also has regular colonoscopies, skin checks, breast exams and blood tests. She doesn鈥檛 mind feeling like she鈥檚 on a medical treadmill because it makes her feel like she is in control. 鈥淚 always get anxious in the lead-up to each test and think they鈥檙e going to find something,鈥� she says. 鈥淏ut then I get screened and told everything鈥檚 OK and I can have a break from the GP.鈥�

Coutts recently had her 6-year-old son tested for TP53 mutations and found out he鈥檚 not a carrier. 鈥淚t was the best news ever,鈥� she says. Her 21-year-old daughter is still deciding whether to take the test. 鈥淚 want it to be her choice because it鈥檚 such a big thing to know. Finding out at that age could completely change your outlook on life.鈥�

The Garvan trial has now confirmed that whole-body MRI is effective at spotting cancers in people with Li-Fraumeni syndrome while the tumours are still at an early, curable stage 鈥� something Coutts finds reassuring for her and her family.

鈥淚 still have a letter from years ago from a doctor who told me there was no guarantee that screening would help me or extend my life,鈥� she says. 鈥淏ut now it鈥檚 been proven to work and it鈥檚 a huge help 鈥� I wouldn鈥檛 have it any other way.鈥�

Combining genomic screening with whole-body scans has identified 16 hidden tumours in 30 people. The technique should help identify curable cancers in people who are genetically at-risk and save lives.

Advances in reading and understanding the genome have made it easier to identify people who are particularly susceptible to cancer. For example, people with mutations in the BRCA1 and BRCA2 genes are more likely to get breast cancer. But often this information isn鈥檛 particularly helpful. While some women with BRCA mutations opt for pre-emptive breast removal, there is little that can be done about some other genetic susceptibilities.

One of these, caused by mutations in a gene called tumour protein p53 (TP53), is present in around 1 in 10,000 people. It greatly increases the chances of getting a range of cancers, including in the breast and brain. About half the people who have such mutations develop a cancer by the age of 30.

鈥淕enomics is revealing听patterns of mutations听that increase cancer susceptibility anywhere in the body, not just in one particular organ, but we haven鈥檛 known what to do with this information because how do you screen everything?鈥� says 听at the Garvan Institute of Medical Research in Sydney. 鈥淭hat鈥檚 where whole-body MRI could help.鈥�

A team led by Thomas and his colleague 听has found that such screening of people at genetic risk can detect hidden tumours before they turn deadly.

Curable tumours

The team performed head-to-toe MRI scans on 30 people aged between 18 and 62 with TP53 mutations. The scans detected three new, early-stage cancers, as well as two recurrences of previous cancers and 11 benign tumours.

Of the three new cancers, two were in the prostate and one was in the lower back, and all were still at a curable stage.

The researchers then analysed data from 578 people from six countries who have TP53 mutations and have had whole-body MRI scans. In total, they found that the technique had identified 35 curable, early-stage cancers, including in the breast, bowel, bone, prostate, lung, kidney and thyroid.

鈥淎s you can imagine, it鈥檚 an enormous relief for these people,鈥� says Ballinger. Since the scanning study began in 2012, all but one of the 30 participants have chosen to continue getting annual, whole-body scans.

The early-onset cancer caused by TP53 mutations is called Li-Fraumeni syndrome. These mutations can be passed down from a person鈥檚 parents or can arise spontaneously in embryos.

鈥淭here鈥檚 extreme stress for families with Li-Fraumeni syndrome,鈥� says Thomas. 鈥淚magine being 20 years old, one of your parents has died of cancer, and your older siblings now have cancer and you鈥檙e wondering when you鈥檙e next.鈥�

Natalie Coutts was one of the participants found to have benign tumours, in her uterus and kidneys, and says she is glad they were identified and removed. 鈥淚 want every lump and bump investigated because there鈥檚 such a high risk of cancer,鈥� she says. 鈥淚 don鈥檛 care how many times I鈥檓 under the knife because I have children and I want to be around for them.鈥�

Family risk

As well as benefiting people with Li-Fraumeni syndrome, regular whole-body MRI scans may help people with other susceptibility genes, like BRCA2, says Thomas. While BRCA2 mutations are typically associated with breast and ovarian cancer, they also seem to be linked to other cancers like sarcomas. 鈥淚f you just do breast screening, you might miss a curable sarcoma in a person鈥檚 thigh,鈥� Thomas says.

Although whole-body MRI is a promising tool for picking up new tumours in people with clear cancer risks, we still need to show that it leads to improved survival, says at the QIMR Berghofer Medical Research Institute in Brisbane.

Testing the genomes of the whole population would be unwise at this stage, Hayward says. 鈥淢ainly because you鈥檇 find numerous mutations that may or may not increase cancer risk and that could cause unnecessary anxiety,鈥� he says.

For the time being, it is better to reserve genetic screening for people with family or personal histories of cancer, Hayward says. 鈥淚f someone has a family history of breast cancer, it might be useful to test for the BRCA mutations, but I don鈥檛 think we should be looking for mutations in every gene in everyone right now.鈥�

Journal references:听JAMA Oncology, DOIs: and

Read more: Counting genetic mutations predicts how soon you鈥檒l get cancer

People who have had amputations can control a virtual avatar using their imagination alone, thanks to a system that uses a brain scanner.

, which translate neuron activity into computer signals, have been advancing rapidly, raising hopes that such technology can help people overcome disabilities such as paralysis or lost limbs. But it has been unclear how well this might work for people who have had limbs removed some time ago, as the brain areas that previously controlled these may become less active or repurposed for other uses over time.

at IDC Herzliya, in Israel, and colleagues have developed a system that uses an fMRI brain scanner to read the brain signals associated with imagining a movement. To see if it can work a while after someone has had a limb removed, they recruited three volunteers who had had an arm removed between 18 months and two years earlier, and four people who have not had an amputation.

While lying in the fMRI scanner, the volunteers were shown an avatar on a screen with a path ahead of it, and instructed to move the avatar along this path by imagining moving their feet to move forward, or their hands to turn left or right. The people who had had arm amputations were able to do this just as well with their missing hand as they were with their intact hand. Their overall performance on the task was almost as good as of those people who had not had an amputation.

鈥淎lthough the amputees鈥� performance is a little bit behind the control group, the big picture shows they are almost the same level, and still using the missing arm in their brain,鈥� says Cohen, who presented the research at the IEEE EMBS Conference on Neural Engineering last month.

Because the system requires a person to be inside a brain scanner, it would not be possible to use it outside a lab. But Cohen thinks that a new technology called functional near infrared spectroscopy will make it possible to read the same brain signals with portable devices. This may lead to new ways for people who have had limbs removed to control prosthetic devices.

But Dario Farina, of Imperial College London, doesn鈥檛 think such a system is likely to be very useful for amputees. 鈥淭here are alternative techniques that are far superior for prosthetic control,鈥� he says.

The fMRI interface only distinguished four commands: forward, stop, left and right. Prosthetic controllers that work by detecting muscle signals at the stump of the severed limb can distinguish more commands, respond more quickly, and allow the user to control the force or speed.

Farina thinks Cohen鈥檚 system could be more useful for locked-in patients, who have no means of communicating except via brain signals. 鈥淔or other types of patients, this is a good performance, which is promising,鈥� he says.

Read more: Amputees control virtual prosthetic arm using nerve signals

听

ARTIFICIAL intelligence is taking image recognition tips from a real expert: the human brain.

Using fMRI brain activity scans as a training tool has boosted the ability of machine learning algorithms to recognise objects. The technique could improve face recognition systems or help autonomous vehicles better understand their surroundings.

Machine learning is still a long way behind humans when it comes to tasks like object recognition, says . So his group trained algorithms to process images more like we do.

They analysed how regions of the brain鈥檚 visual cortex responded to images containing four different types of object: humans, animals, buildings and food. The data came who viewed more than 1200 images while an fMRI machine measured their brain response.

The different objects had their own corresponding pattern of brain activity, and the strength of the signals indicated how difficult each image was to classify.

The team used this information to train its machine learning algorithms. If an algorithm made a mistake on an 鈥渆asy鈥� image, it was more heavily penalised than if it made erred on a 鈥渄ifficult鈥� image. This feedback essentially told the system what information it should base its classifications on to minimise errors. As a result, it performed better on images easily recognised by the brain, effectively making decisions in a more human-like way ().

鈥淚f an AI makes mistakes that a human would make, humans will continue to trust that system鈥�

Training a basic image classifier with the fMRI data improved its accuracy by 10 to 30 per cent across the different categories, says team member at the University of Notre Dame, Indiana. The method could allow relatively basic machine learning models to approach the accuracy of state-of-the-art neural networks, he says.

Algorithms that make decisions in a similar way to us could also be easier to understand and trust, says Cox. Computer systems sometimes make mistakes that humans wouldn鈥檛 鈥� like Tesla鈥檚 Autopilot system failing to notice a white trailer against a bright sky. Systems trained on brain data would make mistakes in a more human way. 鈥淎nd if you make mistakes that a human would make, humans will continue to trust that system,鈥� says Cox.

鈥淭hese are preliminary results, but they鈥檙e impressive and they suggest that this is a new line of work that could be really fruitful,鈥� says at the Medical University of South Carolina in Charleston.

Next, the researchers will look at how cells in rat brains react to different images. The hope is that understanding how the brain works at this level could lead to systems that more closely mimic human decision-making. 鈥淭he really exciting stuff is going to come from looking at the fine-grain detail of how individual cells are connected and how they鈥檙e firing,鈥� says Cox.

This article appeared in print under the headline 鈥淏rain scans aid machines to see more like us鈥�

Keep on running. Ultramarathons do take a toll on a runner鈥檚 body, breaking down cartilage and even shrinking the brain, but athletes seem to bounce back from both, with joints repairing before the race has even finished.

For many, marathons represent the ultimate fitness challenge. But a minority go much further, taking part in ultramarathons like the , covering 4500 kilometres from the south of Italy to Norway in 64 days.

This race equates to around 100 marathons, with no rest days, prompting researchers to wonder what such long periods of intense exercise would do to the body.

Uwe Sch眉tz at the University Hospital of Ulm in Germany and his colleagues have spent the last six years finding out. In 2009, they followed a group of 44 runners as they ran the nine-week race across Europe. The team took a portable MRI scanner with them, and periodically scanned the legs, feet, heart, brains and cardiovascular systems of the athletes, as well as taking blood and urine samples.

Scanning feet and leg joints every 900 kilometres, Sch眉tz and his team measured the amount of water that was released from the shock-absorbing cartilage between the bones 鈥� a sign of whether cartilage is breaking down. They found that the runners鈥� cartilage seemed to degrade during the first 2500 km of the race.

But after that distance 鈥� around 60 marathons 鈥� the cartilage seemed to recover, says Sch眉tz, who presented the findings at the in Chicago this week. 鈥淚t was thought that cartilage could only regenerate during rest,鈥� he says. 鈥淲e have shown for the first time that it can regenerate during running.鈥�

Brain size

A runner鈥檚 joints aren鈥檛 the only parts of their body affected. Earlier analyses of the same runners revealed that their brains seemed to temporarily shrink in size by 6 per cent over the course of the race.

The loss may simply be the result of extreme fatigue and undernourishment, but Sch眉tz thinks it could be caused by lack of stimulation. One of the four brain regions that seems to be particularly affected is known to be involved in visual processing. That area may have been massively under-stimulated by 64 days of viewing little other than roads, he says.

Others have suggested that athletes鈥� brains may to divert energy to regions involved in motivation. 鈥淚t is hard to explain what鈥檚 going on,鈥� says Sch眉tz. 鈥淏ut we do see total recovery after six months.鈥�

Sch眉tz says people who run normal marathons won鈥檛 experience the same degradation.

Aerobic exercise is generally beneficial for the brain, helping to stave off depression and dementia.

(Image credit: Helmut Dietz)

The trick could help us worm our way towards better scanners (Image: Monty Rakusen/Getty)

It鈥檚 a new magician鈥檚 trick, sawing a magnetic field in half 鈥� and all you need is a wormhole. Magnetic fields entering it emerge at the other end as if teleported through space. The feat has practical applications too, such as improving MRI scans.

Despite its evocative name, the wormhole is no portal in space-time, but it does allow a magnetic field to disappear in one region and then re-emerge unchanged elsewhere. Alvaro Sanchez at the Autonomous University of Barcelona, Spain, and his colleagues were inspired to create one by a theoretical .

鈥淭hey modified our earlier mathematical constructions in a very clever way so that an artificial wormhole could be built using present engineering techniques,鈥� says Matti Lassas at the Helsinki University of Technology in Finland, who co-authored the earlier paper.

Sanchez鈥檚 team had transferred a magnetic field across space with a length of special tubing that acted as if it were a hose that could carry magnetic fields without them losing strength. But external magnetic fields would be able to distort the fields inside the hose. To look like a wormhole, the tube itself had to be made invisible.

鈥淲e needed to make a 3D magnetic cloak to hide the magnetic hose,鈥� Sanchez says. To do that, they used metamaterials 鈥� artificial materials that interact in unusual ways with electromagnetic fields and that may some day be deployed to build invisibility cloaks for light.

(Image: Jordi Prat-Camps and Universitat Aut貌noma de Barcelona)

The team nested the hose inside a sphere of superconducting strips that deflect incoming fields (silver layer in figure above). But that deflection would be detectable, so they placed another sphere, this time of magnetic material, inside the strips to hide the superconductors (gold interior layer).

鈥淲e have a very fine-tuned concentration of attraction and repulsion,鈥� Sanchez says. 鈥淭he whole object is magnetically invisible because of this cancellation.鈥�

The magnetic field, when it pops out of the other end, looks like a magnetic monopole (below) 鈥� something not so far detected in nature.

(Image: Jordi Prat-Camps and Universitat Aut貌noma de Barcelona)

Being able to cloak magnetic fields as they travel across space could help build better MRI scanners, Sanchez says. Wormholes could let multiple magnetic imagers work together without interfering with each other, or could be used to put some distance between bulky sensors and patients 鈥� all without changing the background magnetic field MRIs rely on. His team is planning to reach out to local doctors to guide future research.

The magnetic wormhole is an impressive demonstration of the power of metamaterials, Lassas says. 鈥淚t makes the scientific work on invisibility cloaking a step 鈥� or in fact, a leap 鈥� closer to real life applications.鈥�

Journal reference: