DSPs in Medical Imaging
Just as they have played a key role in advancing industries such as PCs, cellular telecommunications and automotive, digital signal processors (DSPs) are playing an increasingly important role in advancing medical imaging to provide faster, more accurate diagnoses and treatment.
The evolution of medical imaging
Medical imaging is continually evolving and advancing to improve patient care. The advances that have been made in recent years have been dramatic. For example, X-rays have migrated from film to digital and images can be taken and displayed so rapidly that digital X-ray machines are now used in surgical procedures (figure 1). Ultrasound machines have steadily become more compact, with cart-based systems increasingly complemented by portable and even handheld units. These portable ultrasound machines are practical for a wider variety of applications, including bringing health care to rural and remote areas, disaster sites, patient rooms in hospitals, assisted-living facilities and even ambulances.
Despite these advances, the medical imaging industry has not evolved as quickly as many other industries where semiconductors have played a key role as the underlying technology. For example, take the mobile phone industry. Over the past 15 years, cell phones have evolved from brick-sized, voice-only devices to units smaller than a deck of cards capable of TV-quality video and broadband faster than many DSL modems, while being affordable enough that most family members have their very own. In fact, today’s smart phones have faster processors and more memory than a state of the art PC fifteen years ago.
So why have cell phones and computers advanced so much faster than medical technology over this period? Part of the reason is because technology companies have focused their attention on these areas and helped advance the technology through investment, research and development. In the meantime, medical imaging companies have often been left without dedicated support, typically supported by vendors with very little understanding of the medical imaging use case.
The amount of attention that medical imaging companies get is poised to increase dramatically with semiconductor companies, like Texas Instruments, making medical imaging a major focus. And a key technology component that will improve medical imaging is the DSP.
The role of the DSP: past and future
DSPs are embedded processors that are ideal for much of the processing required in medical imaging applications. For instance, in digital X-ray, the real-time processing nature of DSPs has been leveraged to increase the speed at which digital x-ray images can be taken and displayed, enabling x-ray imaging to be useful in cardiac and vascular surgery. DSPs have played an even more prominent role in ultrasound’s evolution into a high quality portable imaging machine (figure 2). The combination of high performance and power efficiency offered by a DSP has allowed manufacturers to develop portable ultrasounds that produce quality images while minimizing product size and maximizing battery life.
Today’s research will improve health care even further over the next decade, with DSP-based imaging continuing to play a key role. DSPs will not only help revolutionize the well-known modalities such as X-Ray and MRI, but will enable completely new medical applications that border on the edge of science fiction.
One example of this is from an Israeli company, CNOGA , whose technology uses a video camera to non-invasively measure vital signs such as blood pressure, pulse rate, blood oxygen levels and blood carbon dioxide levels simply by focusing on the person’s skin. Future versions of the technology may be able to identify biomarkers for diseases such as cancer and chronic obstructive pulmonary disease, as well as recommend the proper makeup to go with a person’s skin complexion.
Another application is high-intensity focused ultrasound (HIFU). This technology is part of a trend in health care to reduce the impact of procedures by minimizing incision size, recovery time, hospital stays and infection risk. In the case of HIFU, this involves turning a previously invasive procedure into a non-invasive procedure. There are many medical applications for HIFU, including using the focused ultrasound waves to cauterize bleeding, destroy a cancerous tumor without damaging surrounding tissue, and even melting fat. HIFU has the capability of turning ultrasound, previously a diagnostic modality, into a therapeutic modularity that can be used for everything from saving lives on the battlefield to improving one’s body with cosmetic procedures.
Embedded processors are playing a key role in many of these. Unlike ASICs and FPGAs, embedded processors are inherently flexible, programmable devices, so they present an efficient development platform for rapid development and implementation of sophisticated, adaptive medical imaging algorithms.
Advantages of DSPs for medical imaging
Embedded processors have three key traits often required in demanding, critical applications such as medical imaging:
? They perform in real time. This means immediate bootup, no delays in displaying images, and rapid response after changing functions or parameters.
? They are highly reliable. This includes components that don’t fail even after years of rough handling in the field, the ability to recover quickly from failures due to software bugs, and immunity to electromagnetic interference. Reliability directly affects the system’s total cost of ownership (TCO) for end users such as hospitals and physician groups. By the same token, reliability and TCO also directly affect the brand reputation and competitive position of imaging system vendors.
? They are energy-efficient, compact and affordable. All of these directly affect the level of patient care. Energy-efficient and compact enable portable imaging in places such as first responder, trauma/triage and even home health. Affordability means that state-of-the-art care can be extended to more people, such as residents of developing countries.
Embedded processors also allow end users of the technology, such as hospitals and physician groups, to capitalize on new software and algorithms that universities, programmers and R&D centers are developing to improve diagnostic image clarity, depth and functionality. For clinicians and patients, improved images equal better diagnoses and care.
Finally, equipment vendors also benefit because the software programmable architecture of DSPs enable the ability to field-upgrade machines, which makes their products more attractive to potential customers by providing a certain level of future-proofing, as well as improving machine functionality.
It is no small task developing DSPs that have the performance, reliability and efficiency necessary for medical imaging applications. Yet many of the underlying challenges – including designing for high reliability, high performance and low power – are ones that Texas Instruments has been tackling for decades in industries such as telecom, military communications, and automotive. That experience has produced expertise that can now be brought to bear in medical imaging. By enabling advances in medical imaging, DSPs are making it more practical and cost-effective to bring better, faster and more accurate health care to patients.