Page 553 – Physics World

Home » Page 553

Dosimetry of small fields: the physics behind it

Want to take part in this webinar?

The first part of this webinar focuses on the theoretical aspects of small field dosimetry, whereas the second part – to be held 12 November – presents the best practice in small field dosimetry based on TRS 483 protocol.

This meeting has applied to CAMPEP for approval of one MPCEC hour, and to EBAMP for approval for one CME credit.

The participants of this webinar, presented by Dr Lutz Müller and Dr Hui Khee Looe, will get more insight into the following topics:

What is special about small radiation fields?
Perturbation effects – the interaction between detector and small photon fields.
Strengths and limitations of specific detectors.
Correction factors in SF protocols.
Application of codes of practice for small field dosimetry, mainly IAEA/AAPM TRS 483.
Possible pitfalls in small field dosimetry.

Want to take part in this webinar?

Join the audience

Lutz Müller (left) holds a PhD in nuclear physics from Technical University Munich. After his PhD, he worked in experimental nuclear physics research at the National Nuclear Physics Institute in Padua, Italy. In 2000 he joined IBA Dosimetry and is currently Director of IBA’s International Competence Center (ICC). Hui Khee Looe (right) is Deputy Head of the Department of Medical Physics, University Clinic for Medical Radiation Physics and Radiation Protection, Pius Hospital, Oldenburg, Germany, and holds a PhD from Oldenburg University. He leads a research group on Computational Methods in Dosimetry and has a strong interest in Small Field Dosimetry.

Spin valve uses coupled quantum dots and tiny magnetic fields

Researchers in Switzerland and Italy have developed a method for generating currents of electrons with a known quantum spin without the need for large external magnetic fields. This could enable the development of devices that are compatible with superconducting electronic elements, paving the way for the next generation of highly efficient electronics.

Following the discovery of giant magnetoresistance as well as the observation of spin injection and detection in metals in the late 1980s, a field of research known as “spintronics” emerged dedicated to creating practical devices that exploit electron spin. Semiconductor-based spintronics systems have garnered particular research interest because semiconductors can be integrated within modern-day electronics, thus improving the efficiency and storage capacity of devices. But in order to make useful spintronics devices, researchers need to be able to control and detect the spin state of electrons with a high level of accuracy.

Controlling electron spin

One method for controlling the electron spin current is a device known as the “spin valve”, which usually consists of a non-magnetic material sandwiched between ferromagnetic materials. This material configuration allows electrons with one spin to propagate through the device, while the opposite spin is reflected or scattered away. This occurs because spin propagation depends on the alignment of the magnetic moments in the ferromagnet. Thus, a “spin polarized current” is produced. This is a flow of electrons that, in theory, all are in a set spin state (all spin-up or all spin-down).

Spin-valve trio: Andreas Baumgartner (left), Arunav Bordoloi (centre) and Christian Schonenberger. (Courtesy: University of Basel Department of Physics)

However, these types of spin valve are either not very efficient or require very large polarizing magnetic fields both imposing severe limitations on experiments — for example, experiments involving material systems that are sensitive to magnetic fields. To overcome this and achieve a highly spin-polarized current, researchers are looking for alternative methods to create spin valves using semiconductor materials.

Tiny magnetic fields

Now, physicists at the University of Basel along with collaborators at the National Enterprise for nanoScience and nanoTechnology have created a device that can control electron spin currents without the need for large external magnetic fields and with a high efficiency. In a recent paper published in Communications Physics, they describe how a pair of coupled quantum dots formed in an indium arsenide (InAs) nanowire with nearby individual nanomagnets can be used as a spin valve with an electrically tuneable spin polarization of up to ±80%.

Electric fields control spintronics devices

The team created the quantum dots by electrically defining two areas where electrons in the nanowire are confined in all three spatial directions. They then employed ferromagnetic side gates to generate small local magnetic fields across each dot. This gate-based configuration means that only very small magnetic field of up to 40 mT are needed to obtain a very high efficiency.

The device operates by generating a spin-polarized tunnel current using the first dot, which is then detected by the second dot. By magnetizing the ferromagnetic split gates in parallel or anti-parallel, the researchers can decide whether electrons of a certain spin can pass through each part of the device. The probability that an electron with that spin tunnels through both dots can be controlled using the ferromagnetic side gates, allowing a spin-polarized current to flow when they are aligned but no current at all if they are anti-parallel. The researchers were able to “tune” the device by experimenting with different applied fields and gate voltages. They were able to achieve a high spin polarization efficiency with the potential to reach the theoretical limit of 100%.

New quantum technologies

This type of spin valve could be very useful in applications for which magnetic fields can have a drastic impact on the material characteristics – such as suppressing superconductivity or altering electronic band structures. The manipulation of electron spin with such small magnetic fields may allow researchers to develop new quantum technologies that utilize spin-based quantum phenomena such as entanglement and the confirmation of Majorana fermions in topological superconductors along with facilitating the investigation into new unexplored physics.

Physics in the pandemic: ‘I took it upon myself to keep my classes alive, even if lessons had to be conducted via Zoom’

As a physics teacher, I am used to standing in front of students in a classroom, teaching and engaging them face-to-face. Now that the pandemic has forced schools and universities to reconsider face-to-face classes, two 21st-century concepts have become buzzwords: work from home (WFH) and online distance learning (ODL).

As the popularity of WFH and ODL increased, the popularity of Zoom, the virtual meeting app, increased as well. Suddenly, teachers all over the world found themselves presenting lessons in front of a camera hooked to a computer, “facing” students sitting behind screens many kilometres away.

This method of communication projects the complex physical self into a composed, quiet, virtual identity. For a full week in April, I attended daily four-hour Zoom meetings. During this time, my default device setting was a muted microphone. I turned off my camera for two reasons: to trade upload speed for download speed, and to save every precious bit of my tight data allowance. I was there to listen, so I spoke only when acknowledged, my physical self muted and turned into a meticulously handpicked profile picture. From a fun-loving jokester, I was turning into a faceless, lifeless and soulless being. In short, I was becoming a Zoombie.

Birth of the Zoombies

It is a cliché to say that technology gifts us connectivity, especially at a time when physical connections are impossible. But the trade-off is a lack of physical interaction, and even a reduction in the feeling of being alive. All I saw on my screen or heard from my speakers was a stream of ones and zeroes; all sights and sounds were converted to electrical currents. I sat in front of the laptop with my ears, eyes and brain glued to the device – not with eager focus, but with the sort of lazy attention manifested by hunched shoulders, drooped eyes, and uninspired fingers.

At first, I thought I might be the only one who felt this way. But as my Zoom meetings increased in number and frequency, I noticed that most participants were doing the same thing. Microphones muted, cameras off, sporadic chat messages: lifeless personae drudging in and out of meetings. Soon, I realized that students may be having the same experience with ODL delivered in synchronous sessions. Muted, unresponsive, answering only when asked, detached and bereft of life, the numbers of Zoombies were increasing at a rate that would put 10 seasons of The Walking Dead to shame. And I would not have any of it.

The fight to stay alive

As a classroom teacher I took it upon myself to keep my classes alive even if lessons had to be conducted via Zoom. By harnessing my more-than-a-decade experience as an educator, I knew I could find ways to bring my virtual classroom to life.

The first conclusion I drew is that synchronous class meetings must be as succinct and targeted as possible to shorten the highly regulated environment of a Zoom meeting. This mode should be reserved for important and urgent matters, such as setting learning targets and deliverables for the week, explaining the most challenging concepts, deciding on a particular class issue, and – I believe this is of the highest importance – providing life and socialization to the virtual classroom.

My second conclusion is that there should be a checking-in routine. Letting students interact with each other, even for a brief period, can inspire attention and engagement. Assigning different hosts and “passing” the microphone whenever possible – all while encouraging candidness and candour in front of the camera, just as you would in a classroom – can also improve interaction.

Third, teachers should seek ways to elicit discussion and interaction among students. This can include integrating live polls or quizzes, asking thought-provoking questions, or demonstrating discrepant events (a favourite of science teachers) in front of the camera. In short, we must inject life into our virtual classrooms. The more the students feel alive, the less their chances of turning into Zoombies.

Fourth, as teachers, we should remember that some of our students may have limited connectivity in terms of Internet speeds and data caps. We must wear the hat of a connoisseur (choosing the best learning material), and a museum curator (laying out the virtual class like a museum with themed galleries) when designing our online courses. We should make each minute count and every moment meaningful. This not only enables students to use their time and data allowances efficiently, it also encourages their agency as learners. Let them explore the virtual classroom on their own in an asynchronous mode, moving and learning at their own pace rather than being herded or forced to follow others.

Fifth, and last, it is lovely to bid students goodbye before they exit at the end of a week or a learning stage, through a synchronous interaction. Once again, this is a way to inject life and make them look forward to the next learning session.

World War Z(oom)

As we try to keep our students from becoming Zoombies, we must also consider the pandemic’s side effect: a general feeling of anxiety. Some of our students might be anxious about missing school and friends, studying at home without a conducive environment or schedule, or worried about family members whose jobs or health have been affected. Leaving our students in such a worried state adds momentum to their Zoombie transformation.

The best weapon we can use in this war against Zoombies is to encourage positive student–student and teacher–student relationships in our virtual classes. For example, how can an online course evoke a genuine interest in everyone’s wellbeing and safety during the pandemic? Is it possible to elicit the contexts in which students are learning, and then use those contexts in designing classroom activities?

These actions may require extra effort, but if we wish to design our online classes in a truly human-centred way, such questions must be front and centre in our considerations. Instead of students feeling that they are listening to pre-recorded lectures, reading plain texts, or taking tests that are automatically checked and marked by an algorithm, a virtual classroom with a “beating heart” makes a potent weapon against Zoombification.

Finally, we need to remember that we are not teaching remotely in normal circumstances, but during a health emergency of global scale. When we accept this truth, it becomes easier to keep students feeling alive in our classes. Because if Zoom turns our students into lifeless and unresponsive beings, that would be scarier than all the world’s zombie movies combined.

A longer version of this article appears as a Letter to the Editor in the journal Physics Education, which – like Physics World – is published by IOP Publishing.

Electronic skin displays human-like reactions to pressure, temperature and pain

Researchers in Australia have designed an electronic skin that displays human-like reactions to pressure, temperature and pain. Madhu Bhaskaran and colleagues at RMIT University developed the material by combining artificial sensors for these three stimuli into a single, biocompatible film. Their design represents a significant advance in our ability to mimic human skin, and could lead to important developments in both healthcare and robotics.

As our largest sensory organ, the skin contains an abundance of sensory neurons that continually monitor the levels of certain stimuli in our surrounding environments. These sophisticated receptors transmit the information they gather to the brain, which makes real-time decisions about how we should react to them. If the levels of any stimuli rise above certain dangerous thresholds, the brain can then trigger reactions that take us out of harm’s way.

Three types of receptor are particularly important for our survival: the Pacinian corpuscle, which monitors pressure; the thermoreceptor for temperature; and the nociceptor for pain. As researchers attempt to mimic the function of our skin in artificial materials, it is crucial for them to recreate the behaviours of these neurons. However, the sheer complexity of their reaction-triggering mechanisms has so far proven extremely challenging to imitate.

Bhaskaran’s team overcame these issues using a device named a “memristor”, which can regulate the current in electrical circuits, while remembering how much charge has previously flowed through it. Just as the brain uses its long-term memory to decide how to react to stimuli, memristors can evaluate when to switch between different memory states, based on stimuli detected by sub-nanometre conductive filaments.

Skin-like sensor — The skin-like sensing prototype device, made with stretchable electronics. (Courtesy: RMIT University)

To develop an artificial skin, the researchers combined a strontium titanate-based memristor with a stretchable, gold-on-silicone (polydimethylsiloxane) pressure sensor, allowing them to mimic the behaviour of the Pacinian corpuscle. In addition, Bhaskaran and colleagues incorporated the memristor into a vanadium oxide temperature trigger, which could be tuned to transition between a metal and an insulator at a defined temperature. This enabled them to imitate the thermoreceptor, as well as four critical functions of the nociceptor.

As well as being transparent, durable and biocompatible, the resulting film exhibited responses to multiple different stimuli that accurately reproduced those of the human nervous system. When applied levels of pressure, temperature and pain rose above human-tolerable thresholds, the sensors became triggered almost instantaneously.

Since the electrical skin is both affordable and easy to manufacture, it opens up new opportunities for advances in healthcare – including the ability to replace damaged receptors with non-invasive skin grafts, and even to augment human experiences of certain stimuli for applications including defence and sports. Elsewhere, it could lead to new advances towards human-like robots, as well as smarter feedback mechanisms for interfaces between humans and machines.

The researchers report their findings in Advanced Intelligent Systems.

The best way to slice a watermelon, how to float yourself with balloons, shining light on whisky

What is the fairest way to slice up a watermelon? A juicy question that you might never have considered unless you are suddenly given the task at a children’s birthday party. Well, physicists in Belgium, France and Italy have now tackled the problem using geometry and calculus. After cutting the whole watermelon in half along its length and then in the middle to yield four equal quarters, the researchers discovered that this “half rule” fell away when then trying to slice the watermelon up into equal thin portions. Instead they found a ”2/3 rule”. So for a spheroid watermelon of length 10 cm, for two slices, the first slice should be made at 3.5 cm along the length but for three slices, the first slice should be 2.1 cm and the second 4.2 cm. After doing the calculations, the researchers tested them on a 4 kg watermelon and used Archimedes principle to confirm the slice volumes were equal. Eureka!

A week or so ago the illusionist David Blaine floated over the Arizona desert while hanging on to a bunch of helium balloons. Now if you want to do this – and I am not recommending it – how many balloons would you need? The physicist Rhett Allain has done all the heavy lifting and provides the answer in “Let’s calculate how many balloons David Blaine needed to float”.

Counterfeit drinks cost the UK economy more than £200 million in lost revenue each year and this is a real concern in Scotland, where whisky production is a multi-billion pound industry. Indeed, rare bottles of whisky can fetch as much as £1 million, so collectors are keen on knowing what is in their bottles – but do not necessarily want to open them to find out.

Now, researchers at Scotland’s University of St Andrews have developed a new laser-based spectroscopy technique that produces a chemical fingerprint of a whisky, without having to open the bottle. Previous attempts at doing this had not been successful because of the strong signal generated by the laser striking the glass. The team get around this by using a ring-shaped laser beam that is tightly focussed on a small region in the liquid. By only gathering the spectroscopic signal from this focal point, they avoided the signal from the glass. You can read more in a paper published in Analytical Methods.

Opportunities and fundamental challenges for anion exchange membrane fuel cells

Want to learn more on this subject?

Watch now

In recent years, advances in alkaline exchange membrane fuel cells (AEMFCs) with anion exchange membrane (AEM) solid polymer electrolytes have gained traction due to their distinct, and potentially game-changing, advantages over proton exchange membrane fuel cells. There has been growing excitement in the past 2–3 years, especially as AEMFCs have reached a stage in their development where state-of-the-art cells are reaching comparable power densities to proton exchange membrane fuel cells (PEMFCs) and are able to operate stably for more than 1000 hours.

However, AEMFCs still need to address at least three critical issues if they are to be deployed in the field:

Water management in AEMFCs is more complex than PEMFCs and there is a tendency for significant water accumulation and flooding at the anode – sacrificing both performance and longevity.
There is a need to reduce costs significantly below the PEMFCs and doing this will require completely PGM-free catalysts.
Management of CO₂ and mitigation of CO₂-related performance losses.

This presentation will begin with an introduction to how AEMFCs operate, their similarities and differences to PEMFCs. Then, state-of-the-art performance and durability will be shown along with an explanation for how these two were systematically improved over the past few years due to advances in materials as well as innovations in electrode design and reactor engineering. Next, the three issues above – water management, PGM-free catalysts and CO₂ management – will be discussed, with a focus on the fundamental thermodynamic, kinetic and transport barriers that remain. Finally, an outlook on the future of the technology and important areas for research will be
discussed.

Want to learn more on this subject?

Watch now

William (Bill) Mustain is a Professor in the Department of Chemical Engineering at the University of South Carolina (USC). In 2017, he moved to USC from the Department of Chemical and Biomolecular Engineering at the University of Connecticut (UConn) where he was an Associate Professor and the United Technologies Corporation Professor of Engineering Innovation. He joined UConn as an Assistant Professor in 2008 and was tenured and promoted to Associate Professor in 2013. Professor Mustain has worked in several areas related to electrochemical energy generation and storage including: high-capacity materials for lithium-ion batteries, catalysts and supports for proton exchange membrane and anion exchange membrane fuel cells and electrolyzers, electrochemical synthesis of fuels, electrochemical control of biological systems, the purposeful use of carbonates in low-temperature electrochemical systems, and the electrochemical capture and utilization of CO₂. He has been the PI or co-PI on approximately $10m of externally funded research projects. He has published more than 100 peer-reviewed articles and three book chapters to date and has more than 100 invited and conference talks. He has been the recipient of several awards including the 2009 Illinois Institute of Technology Young Alumnus Award; 2013 US Department of Energy Early Career Award; 2014 Connecticut Quality Improvement Platinum Award; 2014 Supramaniam Srinivasan Young Investigator Award (awarded by the Energy Technology Division of The Electrochemical Society); 2017 UConn Chemical Engineering Faculty of the Year Award, 2019 USC Chemical Engineering Publication Award; and 2015–2016 Fulbright Scholar Fellowship.

Advanced radiotherapy – at a fraction of the cost

EmpowerRT’s founder and president Sha Chang trains staff at the Cancer Diseases Hospital in Zambia to use EmpowerRT’s software. (Courtesy: Cathy Mwaba, Zambia Cancer Diseases Hospital)

EmpowerRT, a social enterprise company spun out from the University of North Carolina (UNC), has an impressive goal: to help resource-limited cancer clinics in developing countries improve radiotherapy, at just 5–10% of the cost of purchasing modern technology.

Currently, there is a huge disparity in cancer care across the world. About 70% of deaths from cancer occur in low- and middle-income countries (LMICs) and, according to the World Health Organization, it is these LMICs that will face the highest increase in cancer death rates in decades to come.

One underlying reason for this disparity is that many cancer clinics in LMICs still rely on older generation radiotherapy equipment, with no funds available to update to modern digital-based systems and the associated infrastructure. And while advances in radiation therapy technologies, such as the introduction of intensity-modulated radiotherapy (IMRT, which has been standard-of-care for years in developed nations), can improve tumour targeting and clinical outcomes, such systems are simply too expensive for many cancer clinics in LMICs.

“This is the really sad truth,” says EmpowerRT’s founder and president Sha Chang. “Many patients in developing countries die and suffer, not because the world doesn’t know how to treat their cancers, but because they have no access to the treatments. Our mission is to help people in low-resource countries to improve the quality of radiotherapy using existing resources, without spending millions of dollars that they don’t have.”

EmpowerRT is tackling this problem by offering a technology that enables any radiation delivery system to provide IMRT, using a low-cost compensator placed in front of the treatment beam. The compensator comprises a styrofoam mould filled with tungsten granules. It can be fabricated using a milling machine, assembled on site and, after treatment, the tungsten can be recycled to treat other patients. “The core belief of EmpowerRT is that we can advance cancer care by doing more with less, doing more with what is already available, and doing it simply”, says Sha.

Similar to modern radiotherapy systems, the IMRT compensator shape is individually designed to attenuate the radiation beam to maximize dose to the tumour target while minimizing dose to nearby healthy tissue. “Using this approach, we can make existing radiation therapy machines in LMICs deliver a more advanced treatment. That’s the beauty of it,” says Sha. She explains that the recyclable compensator IMRT solution was developed at UNC before multileaf collimator (MLC)-based machines were widely available, and was used in treatments for 14 years and 1400 patients before UNC terminated its use after migrating to digital-based radiotherapy.

“We developed the recyclable compensator IMRT technology at a time when we were in a manual operating environment that’s still used in many low-resource settings. So we know this technology is feasible in that environment,” says Sha. “We want to give our proven technology a second life that can benefit more cancer patients.”

In addition to the recyclable compensator IMRT technology, EmpowerRT also offers treatment planning software and e-chart, a digital patient chart software designed for manually operated clinics. “Our goal is to bring the many benefits of a record and verify system to low-resourced clinics in LMICs,” says Sha. “We are also bringing in grid therapy, a novel therapy that benefits large and late-stage tumours, which are more prevalent in developing countries.”

The company also provides extensive training and support, to help sites introduce IMRT – everything from quality assurance and patient selection to setup – in order to establish an effective and safe treatment programme.

Inaugural installation

EmpowerRT has now installed this technology at its first site – the Cancer Diseases Hospital in Zambia. The country’s sole cancer hospital, it has three basic radiotherapy machines, none of which were previously able to deliver IMRT.

The EmpowerRT team offered remote training on the treatment planning software and also travelled to Zambia three times for extended onsite training and dry runs. Sha notes that the training focused on knowledge transfer: “we train the trainer, and then watch the trainer train others,” she explains.

Radiotherapy training at the Cancer Diseases Hospital in Zambia. (Courtesy: Cathy Mwaba, Zambia Cancer Diseases Hospital)

“The training was excellent and the time spent was more than expected, with various hands-on and ‘real patient’ practice cases,” says Mulape Kanduza, chief medical physicist at the Cancer Diseases Hospital. “My first impression was ‘why is this renowned professor from UNC interested in working with our department?’. Effort was put both by the trainees and trainers and I believe the output is evident.”

Challenges such as machines breaking down and long delays in performing repairs, which are not uncommon in a LMIC environment, exacerbated by COVID-19-related travel restrictions, mean that the Cancer Diseases Hospital has not been able to complete an IMRT patient treatment yet. “But we have overcome many unexpected challenges, and the Zambia team of doctors, physicists and radiation therapy technologists all learned a great deal,” says Sha. “They are confident that they can do this new procedure independently, with us remotely doing the review and monitoring. That’s the success.”

“The future is very positive in that we have created a lasting relation to draw from,” Kanduza tells Physics World. “The positive outcome [will be] seeing an IMRT patient treatment being executed and knowing that the outcomes will be much better than treating with other conventional techniques.”

“Improved quality assurance in the workflow, using the Excel-based e-chart to review treatment records, has allowed us to think differently and to acknowledge change,” Kanduza adds. “One measurable outcome is that EmpowerRT’s e-chart system can work in a low-income setting such as ours.”

Looking ahead, EmpowerRT is now in talks with cancer clinics in Nigeria and Indonesia. The company is also partnering with organizations such as Clinton Health Access Initiative, which has worked for decades to reduce the burden of disease in LMICs, and Rayos Contra Cancer, which focuses on global radiotherapy training.

“We are also writing an NIH small business proposal targeted to feasible and effective cancer solutions in LMICs,” says Sha. “The next step is not only looking for where we can help, but how as a social enterprise company we can continue to grow, to partner with more like-minded global cancer care organizations and make a difference together.”

How the Doomsday Clock ticks, dusting on the Moon, Proxima reflects on women in STEM

For almost 75 years, the Doomsday Clock has monitored how close humankind is to global catastrophe. With the clock now closer to midnight than ever before, the science writer Rachel Brazil talks to Physics World’s Matin Durrani about how the clock is set by the Bulletin of the Atomic Scientists and how physicists can engage in public debates about the challenges facing civilization.

Space travel is one challenge that humans have overcome, and we could soon embark on a second era of exploration that culminates in the colonization of the Moon. Lunar dust, however, could be the nemesis of Moon dwellers because the stuff gets everywhere and can damage spacesuits and other essential technologies. Xu Wang and Ben Farr of the Laboratory for Atmospheric and Space Physics at the University of Colorado Boulder join the podcast to explain how beams of electrons could help keep equipment dust free on the Moon.

The film Proxima looks at the themes of motherhood, separation and sexism from the perspective of an astronaut who is preparing to spend a year on the International Space Station. Physics World’s Tushna Commissariat is on hand to give her impressions of the film and to chat about our interview with the film’s writer and director Alice Winocour.

Stable recordings enable ‘plug-and-play’ control of brain–computer interface

Brain–computer interfaces (BCIs) offer the potential for people with severe motor disabilities to control external assistive devices with their mind. Current BCI systems are limited, however, by the need for daily recalibration of the decoder that converts neural activity into control signals. Researchers at the UC San Francisco Weill Institute for Neurosciences have now employed machine learning to help an individual with tetraplegia learn to control a computer cursor using their brain activity, without requiring extensive daily retraining.

To record neural activity, the team used a 128-channel electrocorticography (ECoG) implant, a pad of electrodes that’s surgically placed on the surface of the brain. Already approved for seizure monitoring in epilepsy patients, ECoG arrays provide greater long-term stability than the pincushion-style arrays of sharp electrodes used in previous BCI studies, which are more sensitive but tend to move or lose signal over time.

“The BCI field has made great progress in recent years, but because existing systems have had to be reset and recalibrated each day, they haven’t been able to tap into the brain’s natural learning processes,” explains senior author Karunesh Ganguly. “It’s like asking someone to learn to ride a bike over and over again from scratch.”

ECoG neural implant — The ECoG implant. (Courtesy: Noah Berger/UCSF)

The researchers developed a BCI algorithm that uses machine learning (based on an adaptative Kalman filter) to map the neural activity recorded by the ECoG electrodes to the user’s desired cursor movements. Over a period of roughly six months (two months after ECoG implantation), they conducted experiments to assess two approaches for adaptation of this decoder algorithm as the user learns control.

First, they tested the existing practice of resetting the algorithm each day. Here, the participant imagined movements of his neck and right wrist while watching a cursor move across the screen, and the algorithm gradually updated itself to match the cursor’s movements to the generated brain activity. This daily initialization, however, could require hours for the participant to master cursor control; some days, he could not achieve control at all.

Next, the researchers examined an approach called long-term closed-loop decoder adaptation (CLDA), in which decoder weights from the previous day’s session are used for the current session, without resetting the algorithm each day. They found that continued interplay between the participant’s brain signals and the algorithm resulted in continuous improvements in performance over many days.

When the team reset the BCI algorithm after several weeks of continuous learning, the participant rapidly re-established the same patterns of neural activity required to control the device – effectively retraining the algorithm to its former state. “Once the user has established an enduring memory of the solution for controlling the interface, there’s no need for resetting,” says Ganguly. “The brain just rapidly convergences back to the same solution.”

The researchers next examined whether long-term CLDA could enable long-term stable performance without the need for retraining or recalibration – the so-called “plug-and-play” performance that could enable broader use of BCIs. They found that once expertise was established, the participant could simply begin using the BCI each day. His performance did not decline over 44 days without retraining, and he showed little performance decline after days without practicing.

At this stage, the participant was also able to “stack” additional skills, such as clicking a virtual button, without impacting his ability to control the cursor. To differentiate clicking from cursor control, the researchers trained a linear support vector machine (SVM). “One of the main innovations was the rate at which the Kalman filter and SVM parameters were adapted across days in order to track user learning and neural plasticity,” Ganguly notes.

Ganguly and colleagues conclude that by leveraging the stability of ECoG interfaces and neural plasticity, this approach offers reliable, stable BCI control. This stability may prove even more important for long-term control of more complex robotic systems such as artificial limbs – a key goal of the next phase of this research.

The team is now in the process of enrolling the next subject, says Ganguly. “We are also developing new approaches to increasing communication rates; this involves more sophisticated algorithms and using a denoised decoding of neural activity. And we are testing control of a six degrees-of-freedom robotic arm – we currently have a simulation in a physics engine,” he tells Physics World.

The results are reported in Nature Biotechnology.

Silk hard drive stores optical data

Researchers in the US and China have made the first silk hard drive using a technique called tip-enhanced near-field infrared nanolithography (TNINL). The device, which can store digital data with a density of 64 GB per square inch, is robust in the face of harsh conditions such as heat, moisture, gamma radiation or high magnetic fields. While a silk-based hard drive is unlikely to match the speed and storage capacity of state-of-the-art solid-state drives at the same cost, its unique set of features makes it promising for electronics that could be implanted in the body.

Lithography techniques are routinely used to make devices with optical storage densities as high as several hundred GB per square inch. However, it can be time-consuming to create small features with these methods, and it also requires costly and sophisticated fabrication procedures. Another drawback is that diffraction limits the resolution of conventional optical lithography to around half the wavelength of the illuminating light – roughly hundreds of nanometres for visible light. This makes further increases in storage density (beyond current industry standards) difficult to achieve.

Overcoming the diffraction limit with near-field light

An alternative technique known as scattering-type scanning near-field optical microscopy (s-SNOM) offers a possible solution. This technique gets around the diffraction limit because it depends on near-field light – that is, the component of light that “wraps around” surfaces – instead of the diffraction-prone far-field light (which propagates away from the structures that scatter it). Near-field light is already commonly used in nanoscience as a tool to fabricate, manipulate and characterize photosensitive structures. For example, researchers use so-called “evanescent” light fields for nanolithography of polymeric materials, and for processing nanoscale features in material surfaces when making optical nanodevices.

Researchers led by Tiger Tao of the Chinese Academy of Sciences in Shanghai, Mengkun Liu of Stony Brook University in New York and Wei Li of the University of Texas at Austin have now used a variation of s-SNOM to write and read information on a layer of silk protein.

To make their silk drives, the researchers took silk extracted from silkworm cocoons and spin-coated a thin film of it (in the form of an aqueous solution) onto either a gold or a silicon substrate. They then created nanoscopic patterns in this film using an s-SNOM equipped with a mid-infrared laser light source. The frequency of the laser could be tuned between 1495 and 1790 cm^–1, which covers the wavelengths at which silk proteins absorb light. By focusing the laser onto the sharp tip of an atomic force microscope placed near the silk surface, they were able to induce topological and/or phase changes in the silk film at roughly 30-nm length scales.

Information can be repeatedly written and erased

With these induced changes spaced 100 nm apart, the team achieved a data storage capacity equivalent to around 64 GB per square inch, Liu says. While the patterns thus created are robust to harsh conditions, the method nevertheless allows information to be repeatedly written onto and erased from the film.

The advantages of the technique are many, Liu tells Physics World. Thin silk films are easy to spin-coat onto various substrates, including curved and soft materials. Such films also offer two entirely different ways to read out the encoded data. The first option is to use straightforward AFM “mapping” in which topographic protrusions are recognized as “1s” while flat topographical features correspond to “0s”. The second, more innovative way is to use a laser readout technique that differs from conventional optical methods in that the spectroscopic information thus obtained is multi-coloured. What is more, by varying the laser’s writing power, the researchers demonstrated that they could achieve so-called “higher-dimensional” or greyscale data storage, where individual data points take on values other than a binary zero or one.

Finally, Liu notes that silk is an organic material that mixes and interfaces well with many biological systems, including biomarkers in blood. The information in these biomarkers could therefore be encoded and stably stored in a silk-based hard drive. Indeed, members of the team, who report their work in Nature Nanotechnology, now plan to integrate their silk drive with biological materials for bio-information storage applications. “We will be investigating body implantation in the near future as well,” Liu says.

Browse articles by content type

Dosimetry of small fields: the physics behind it