Discover some of our current research and its applications.
Medical sensing technology
Professor Esther Rodriguez-Villegas leads a multi-disciplinary team of Postdoctoral researchers and PhD students. The expertise of the group includes high performance analogue circuit design, biomedical signal processing, electronic circuit packaging, and high performance PCB design.
She recently led the GUES team from Imperial College London to win $120,000 in the XPRIZE Nokia Sensing XCHALLENG. The international competition aimed to develop breakthrough high impact medical sensing technologies that will ultimately enable faster diagnoses and easier personal health monitoring. Prof Rodriguez-Villegas and her team of researchers have developed Acupebble, a wearable, wireless device, the approximate size of a pound coin, which sticks onto a person’s neck or chest to detect sounds emanating from the heart and respiratory system. AcuPebble uses advanced algorithms to sift through a range of sounds to determine parameters that may indicate deteriorating health or illness in patients.
A versatile sensor
Based on their research and development work over the past eight years, the Imperial team believes that AcuPebble could be used in a range of clinical settings including as a diagnostic tool for diseases such as sleep apnoea, chronic obstructive pulmonary disease or asthma, a health monitor and as an early warning device for both adults and children. The sensor will collect data in real-time, interpret it automatically and transmit this information to an application that can be downloaded onto any smart device, so that patients can be self-diagnosed and doctors can monitor their patients anywhere in the world. The researchers believe that AcuPebble could be used to improve the diagnosis of a range of respiratory and cardiac conditions, including sleep apnoea, whooping cough, pneumonia, chronic obstructive pulmonary disease and congestive heart failure.
The researchers, in conjunction with academics from the National Hospital for Neurology and Neurosurgery, have also carried out a small pilot clinical study using the AcuPebble. The trial consisted of 30 people, including 20 controls and ten patients who had been referred to the hospital with suspected sleep apnoea. The team found that AcuPebble could automatically detect at least nine out of ten individual apnoea episodes in those included in the pilot study.
AcuPebble was able to automatically analyse a range of acoustic signals from both inside and outside the body to determine the sounds that could indicate sleep apnoea. For example, the sensor analysed turbulence in the airways, which can indicate an obstruction, the depth and duration of breathing and other vital signs such as a person’s heart rhythm. The device was able to provide accurate information about breathing rate, heart rate and lung volumes and the type of apnoea a patient was experiencing.
Professor Rodriguez Villegas is joined in Team GUES by her research assistants Guangwei Chen and Syed Anas Imtiaz, who are all part of the Department of Electrical and Electronic Engineering at Imperial College London.
-Laura Gallagher and Colin Smith, Communications and Public Affairs
Professor George Constantinides is Head of the Circuits and Systems Group in the Department of Electrical and Electronic Engineering. He has formed the Imperial spin-out company Novocore, with Dr Christos Bouganis.
What do you hope to achieve through your research?
Standard computer architectures like those you find in desktop PCs are general purpose and not suited to specific applications. We want to design customised systems that can be tailored to perform specialised tasks, such as real-time face recognition, automatically, without using large amounts of hardware. In embedded systems, which are small specialised systems within larger devices, such as mobile phones, the capacity to engage with apps has increased, but battery life has reduced significantly. There is a market for devices which are considerably more energy efficient.
How will you do this?
The complexity of hardware systems lies in producing the hardware itself. However, by changing the way the hardware is designed, we can ensure the hardware ends up cheap, simple and power efficient. We use FPGAs (field programmable gate arrays), a technology which allows us to reconfigure hardware design, using small low energy platforms. This avoids the need to design a new chip, costing potentially millions of pounds, for every new technology.
Energy efficiency is no longer only a concern for embedded systems and companies like Google, who use computer systems that operate at the highest possible level, must now monitor their energy consumption. We are working with high performance computing specialists to find out how our technologies can be applied in these areas. Through Novocore, we are trying to bridge the gap between research and industry.
— Kailey Nolan, Imperial Innovations
Professor William (Tom) Pike is Professor of Microengineering in the Optical and Semiconductor Devices research group. His recent project focuses on increasing integration within three- dimensional circuits, whereby multiple layers of electronic components are incorporated into a single circuit. If more components were integrated into one circuit, devices such as printed circuit boards used in PCs could be made smaller, meeting the continuous push for miniaturisation.
What have you developed?
We have been using a machine called a deep reactive-ion etcher to carve precise holes through silicon wafers. We inject liquefied metal solder into the holes to produce a conductive pathway through the wafer connecting devices on each side. These vertical connections mean we can produce stacks of devices.
What challenges have you faced?
We are trying to make the holes small enough to produce the interconnects for micromachined Mini circuits inventor’s corner sensors, and getting the solder into them is challenging. We use surface tension, as well as gas pressure, to produce a force strong enough to push the liquid solder through the tiny holes. And to make the solder flow, we have had to work in a controlled atmosphere – replacing the air with nitrogen – to prevent oxidisation.
How is this innovative?
The strength of our invention is melting very small balls of solder and using surface tension to direct their flow. With this, we can take various chips that have gone through different processing routes and stack them together. This has been done before, on an individual level, but using the ion etcher and molten solder gives a more reliable and accessible method that can be mass produced.
— Kailey Nolan, Imperial Innovations
Dr Pantelis Georgiou has developed a Wellcome Trust funded bio-inspired artificial pancreas, which aims to improve the treatment of patients suffering from Type 1 diabetes.
Why have you developed this device?
Type 1 diabetes is an autoimmune disease, in which the beta cells in your pancreas (the cells responsible for sensing your blood glucose and releasing the insulin) get destroyed. Traditional insulin injections solve the problem in the short term, but patients still end up having large glycaemic variability, meaning their blood sugar still fluctuates throughout the day – a leading cause of secondary complications like blindness, heart disease and nerve damage. The bio-inspired artificial pancreas can control the blood sugar continuously throughout the day, helping to constrain glycaemic variability and the resulting medical difficulties.
How does it work?
The artificial pancreas is worn externally and combines a continuous glucose monitor that reads your blood sugar, and an insulin pump that infuses the insulin into the body. The innovative aspect of this is the biologicallyinspired microchip which connects the two; we’ve been able to replicate the beta cells using integrated circuits on a silicon micro-chip. With this we can deliver insulin profiles that would be as expected in a healthy pancreas.
How will you develop this technology?
We’ve validated this technology through a patient simulator and were able to regulate the patient’s blood glucose within target levels 93% of the time. This allowed us to begin human clinical trials, which are currently underway at St Mary’s Campus. The next step is to develop this into a larger scale study that will take this from the clinic to the home.
— Kailey Nolan, Imperial Innovations
Dr Simos Evangelou has been working with Dr Daniele Dini (Mechanical Engineering) to develop a new suspension system for road vehicles.
What are you working on?
The suspension system we’ve developed combines the performance improvements of fully active systems with the safety and economy of passive systems. A passive system is what you would find in most road vehicles. An active system includes components which react to an external stimulus and can alter certain parts of the suspension to react favourably to road conditions.
How is your system different?
Because of the nature of a passive system, car suspension systems must be built to work in most conditions, which can lead to shortcomings such as understeering. An active suspension system can, however, react to external information and correct these shortcomings to deliver improved handling, comfort and safety. We have sought to develop a system that retains most of these benefits, while reducing the extent of the active components and, thus, the complexity, cost and maintenance requirements.
How does it work?
Our system uses an electrical mechanism called an actuator to vary the geometry of the passive element of the suspension system. According to the studies we’ve done, this mechanism can be smaller than in fully active systems, which means there is a lower power requirement. In addition, we can integrate the system with current passive suspension systems. This reduces the cost and complexity of the design and means that the system is fail-safe, which will offer a more attractive proposition to vehicle manufacturers than a fully active system. We are in the process of developing prototypes to demonstrate our results.
— Gavin Reed, Imperial Innovation
Professor Kin K Leung joined Imperial as Professor of Internet Technology, following a 20-year research career in telecommunications at Bell Labs and is now Head of the Communications and Signal Processing Group. His research interests remain in telecommunications and he is working on a project that aims to push the boundaries of small cell technology for large-scale commercial applications.
What is meant by small cell technology?
Small cell technology is a term for cell communication with a range of tens of metres or even less: Wi-Fi, for instance. Femtocells are another example; these are cellular base stations that improve network coverage and capacity in small areas by using low transmission power, therefore reducing network interference.
How will you apply femtocell technology?
We want to apply the use of femtocells to large areas, such as shopping centres. If commercial businesses knew precise customer shopping patterns: where they stay, what they buy, they would be able to tailor their services and products accordingly. We call this concept mobility profiling. We can track customer locations as their mobile phones are connected to the closest femtocell and this tracking precision is enough to obtain valuable information for businesses. The tracking and profiling method allows us to maintain user confidentiality; once an individual’s data is collected, it is processed and absorbed into the profiling model parameters and can be instantly deleted.
How could this be developed further?
There are many potential applications. We could profile mobility between competing businesses or, with information from service providers, we could assess purchasing power against other factors, like age group and postcode. We recently filed our first patent application and plan to set up a spin-out company by the end of this year.
—Kailey Nolan, Imperial Innovations