MRI scan through a human head

Medical systems electronics: new inside stories

Advanced medical imaging instruments are getting better at detecting life-threatening conditions - and treating them; but they are also becoming safer for both technicians and patients.

A recent report by Cancer Research UK found that twice as many people are living a decade or more after diagnosis with cancer compared to the same situation in the 1970s: this is due not only to improved treatments but earlier, more accurate diagnosis, and the technology to achieve this has been continuing to advance apace in recent years.

Electronic diagnostic tools such as magnetic resonance imaging (MRI) and X-ray computed tomography (CT), which uses digital processing to improve on the classic X-ray photograph, have made it easier for doctors to analyse potential tumours within a human body without having to resort to biopsies at an early stage.

Newer technologies, such as terahertz imaging, show potential for finding cancers, as well as other medical problems, while technologies traditionally used purely for diagnosis are now being recruited as potential treatment aids in their own right.

Here E&T focuses on three of these technologies – ultrasound imaging, MRI, and terahertz imaging - and looks at some notable indications of how they are evolving medically and electronically...

Ultrasound: real-time sonic boons

Most people are now fairly familiar with live scan images of unborn babies, even if they need a lot of help in working-out just where the baby's head is. Its use during pregnancy makes ultrasound one of the most widely-employed imaging techniques developed thus far. Ultrasound's big advantage over techniques such as MRI and CT is its real-time nature.

Three-dimensional renderings are helping patients decode the images more easily, although healthcare workers are somewhat divided over their therapeutic benefits – many of them have built-up experience on interpreting 2D slices. Bodily organs such as livers and kidneys can show-up oddly in the 3D scans because of the way internal features reflect ultrasound.

The biggest problems for ultrasound are 'noisy' images, along with issues related to the bulk of conventional instruments. Manufacturers are trying to shrink the electronics to make it possible to increase the number of channels – and either increase the effective resolution or cut the time to build each frame – and make it possible to move instruments into cramped operating theatres to help pinpoint tumours during surgery.

Increasing the number of channels means packing more transducers into the probe heads; but this demands the use of high-quality coaxial cables to avoid the signal degrading too much as it passes from the probe tip to the central computer where the core processing occurs. This adds weight and bulk to the system's overall 'footprint' on the surgery floor.

Companies such as Maxim Integrated are now designing low-power analogue-to-digital converters that consume relatively little power so that they can be packed into the probe head itself.

The converted digital signal can then be relayed to the mainframe and computer processor using cheaper, lighter twisted-pair cabling, similar to that used for standard Ethernet network communications in enterprise networking and elsewhere. Even with digitisation at the probe head, ultrasound can generate copious quantities of data at the sampling rates needed to obey Nyquist's theorem, which determines the minimum rate at which a signal needs to be sampled without losing information. The issue for ultrasound systems is that holding to Nyquist's calculation increases power consumption without necessarily providing better accuracy.

The signals of interest are generally recoverable using lower sampling rate than those implied by the ultrasound frequency itself. Researchers from the Technion (Israel Institute of Technology) refer to the technique of using rates significantly lower than the Nyquist rate as 'xampling', and claim that the technique can reduce the amount of data that needs to be transferred from the probe head by almost eight times.

Xampling works with 'beamforming' – a technique borrowed from the world of military radar, to generate a steerable beam of ultrasound. Digital processing can be used to build-up selective interference signals in the output from an array of transducers that generate ultrasound in the form of a narrow steerable beam.

Because the path of this beam is more predictable than with traditional techniques, the xampling technique is less susceptible to noise. Beamforming is also being used by researchers from the German Fraunhofer Institute in more conventional approaches to ultrasound processing, using the ability to 'steer' multiple beams to create images at a more rapid rate.

Like terahertz, ultrasound may have another future in treatment, by using a side-effect of the vibrations on liquids. OxSonics, an Isis Innovation spin-out from the University of Oxford aims to use ultrasound's ability to generate tiny 'bubbles' in the beam's target to help improve the efficiency of drugs to target tumours.

According to OxSonics, the levels needed to generate these nanoscale are similar to those used just for imaging. As they collapse, the bubbles agitate the fluid around them, which helps to drive drugs deeper into the tumours targeted by the ultrasound beam.

MRI: use of proportionate force

Even healthcare laypersons will be aware that MRI scanners use powerful magnetic fields and radiowaves to form images of the body: the magnets inside MRI machines are very good at sucking-up large metal objects that were not screwed down or simply got pushed too close to the tunnel that normally accomodates a significantly less magnetisable human.

Despite the use of powerful magnets, doctors regard the technique as being much safer for patients than CT scans because it does not use ionising radiation. The magnetic fields may seem intense – on the order of 1T for a healthcare full body scanner, which is about that of a scrapyard magnetic hoist – but the limited exposure times that patients endure means they are seen as more-or-less safe. However, the powerful magnets led to MRI coming very close to getting banned by regulations designed to cut down on electromagnetic induction (EMI) affecting employees. A campaign to get them treated differently managed to stave-off the threat; and the European Union is now working on revised regulations to control the exposure of hospital staff to the fields.

The intense fields are needed to recover information from the atoms inside the body. Elements that contain an odd number of protons will align themselves to the field, but a large proportion can be knocked into a more energetic state by a strong RF pulse.

As the nuclei relax into the lower-energy state, they generate electromagnetic pulses that are picked up by RF receivers. The frequency of each pulse depends, albeit on the scale of parts per million, on the electron state of each atom.

The greater the degree to which the electron cloud shields the nucleus – which is subtly different for the different chemical composition of each molecule – the lower the resonant frequency will be. These small differences are used by the MRI instrument's computer to build up a picture of the various soft tissues in the body.

A key problem with MRI is that it is slower to perform than CT, and is not suitable for all types of imaging – it shows-up soft tissue better, but like the simpler X-ray photographs that preceded CT, that is better for analysing problems with bones, and does not cause problems with metal implants.

Although companies aim to use new forms of superconducting magnet to reduce the size of the MRI scanner itself, researchers from French research institute CEA-Leti and Siemens are now going in the opposite direction. With the Iseult project, they are using supercooled magnets to build a huge cylindrical MRI able to generate a field of more than 11T, and which the team expects will provide images of the brain that are an order of magnitude better than possible today. The magnet will be actually 30 per cent more powerful than those to steer particle beams at Cern's Large Hadron Collider.

To avoid the powerful magnetic field leaking out of the instrument, a second set of counteracting coils needs to be wrapped around the magnet itself. The team expects the magnet to be finished in the summer of 2014, and could generate images by the end of the year; but in fact it might not be the way that future super-powerful MRI instruments are designed.

The trouble with conventional approaches to MRI design is the need for helium cooling. Despite being one of the most abundant gases in the universe, helium is actually becoming harder to obtain on Earth. 

Helium recycling has become more common, but research is now looking at other ways to generate the intense magnetic fields required by MRI. Higher-temperature superconductors, such as magnesium diboride (MgB2), look to be the most promising avenue of attack.

Although MgB2 still needs to be cryocooled, its critical temperature is tens of kelvin higher than the niobium-titanium alloy used in most production MRI scanners.

Researchers such as Bartlomiej Glowacki, Professor of Energy and Materials Science at the University of Cambridge, argue that MRI could become part of the future hydrogen economy, in which superconducting power lines are cooled by pipes of liquefied hydrogen. MRI installations would provide natural storage locations for hydrogen, and work with smaller helium-filled heat exchangers to provide refrigeration.

Another way to attack MRI is to improve its responsiveness. To date, patients have to stay perfectly still to allow a good enough image to be built up over a period of seconds or even minutes; but, as researchers from UC Davis in California found, this is still not good enough to pick-up injuries in delicate, complex joints such as the wrist.

The team developed a way of using the scanner to take one image every half-second, delivering a series of images in a half-minute. Although it does not support real-time motion, it provides more information on injuries around the joint.

The bones themselves cause interference as they move, however – interference that is known in these situations as 'banding artefacts'. These dark bands can obscure the area around the joint as it moves. So the UC Davis research team used dielectric pads laid around the wrist to help steer the magnetic field, and move the artefacts away from the joint itself.

In conclusion, it is clear that none of the three technologies surveyed above will usurp the others as more advances are made; but the fact that these well-established technologies are being developed for further patient benefits is highly encouraging. 

Further information

Recent articles

Info Message

Our sites use cookies to support some functionality, and to collect anonymous user data.

Learn more about IET cookies and how to control them