Medical monitoring devices are moving out of the hospital and into the hands of consumers, driving a shift to a platform approach for designing medical electronics. This article looks at areas manufacturers need to be conscious of when considering the appearance and features of their finished product.
By Charlie Jenkins
AT A GLANCE
Medical monitoring devices that were once found only in hospitals and ambulances are now making their way into homes and businesses. This move closer to the consumer requires that medical designs become more portable, more intelligent, and offer robust communications capability. Market success will also require more cost-effective design approaches, beginning with platform-based design.
The current trend toward consumer medical monitoring equipment began with the advent of portable monitoring devices in hospitals. These wearable sensor modules allowed patients to roam freely within the hospital, getting needed exercise to accelerate recovery, while the device continually relayed critical parameters to a central monitoring station. That freedom of movement stimulated interest in devices that could be used in home or work settings, such as oxygen and blood pressure sensors. Making the jump from the hospital to home, however, has added significant design challenges in the areas of portability and device intelligence.
The need for portability implies smaller, more compact designs. It also means a reduced battery footprint, which in turn drives electronics to use as little power as possible. Experience with other consumer devices such as cell phones, audio players, and laptop computers has shown that long battery life is an important factor for consumer satisfaction in portable equipment. Devices with replaceable batteries need to operate for months between replacements. Devices with a shorter operating life must offer rechargeable batteries.
Making a battery system rechargeable creates an additional set of design challenges. Today’s high-capacity rechargeable batteries, including nickel-metal-hydride and lithium-ion types, are complicated to charge. Without careful control of the charging circuit’s voltage and current outputs, such batteries could overheat and possibly explode. An intelligent battery system can contribute as much as one-third of the total device cost. Further the discrete circuitry that provides this type of active battery management can be more costly than the battery pack itself.
Portability may also impose constraints on system connectivity. These devices need wireless telemetry – the capability to relay data to another system for analysis or recording. In a hospital setting, medical devices are able to use proprietary channels in the ISM (industrial, scientific, and medical) radio bands. The proprietary nature of these wireless channels makes it easier to meet quality-of-service telemetry requirements. On the other hand, cost sensitive consumer devices need to operate regionally, nationally, and even worldwide in shared wireless environments (i.e., 802.11x) and are much more susceptible to local interference issues. Wireless connectivity for consumer medical devices must therefore have flexible designs able to be adjusted for each environment.
Device Intelligence Required
Both complex battery management and flexible connectivity requirements indicate that intelligence needs to be built into these consumer medical devices. User interface requirements add to that indication. Consumer medical devices must not depend on the presence of a trained user to initiate operations and interpret results. Devices must be designed to be as simple as possible for the user, which means it must have its own decision-making capabilities. For example, automated external defibrillator devices (AEDs) are becoming popular for public use at stadiums, transportation centers, and commercial businesses. Conventional defibrillators display cardiac signals and depend on a trained user to set energy levels before delivering any required shocks. Consumer-based AEDs must safely and reliably make an interpretation of the cardiac signals to determine whether or not to deliver a shock and at what energy level.
Delivering these kinds of capabilities in a medical device will require a shift in design approach for many development teams. The traditional design style involves creating unique systems by assembling a collection of CPUs, application-specific standard products (such as battery management devices), and interface devices, then hooking them together with programmable “glue” logic. The alternative approach used in very high-volume consumer devices, a custom IC (integrated circuit), is not practical for medical devices because of their relatively low production volume and modest (typically under $2,000) price points.One problem with the traditional design approach is it takes considerable design effort to integrate such a diverse collection of components into a smoothly-operating whole. It also imposes a requirement for the availability of individual components over five to ten years, the typical market life of medical devices. The pace of change in the semiconductor industry makes such component lifetimes increasingly unlikely, therefore designs created using this style will need periodic redesigns to stay in production, as components become obsolete.
The traditional design approach also suffers from the high cost of regulatory certification. When development teams create “point-solution” designs, each must undergo a full certification process. Further, the design changes from component obsolescence forces periodic recertification. Hardware changes made to upgrade a design or to address problems found in the field also force recertification.
 |
| Figure 1: FPGA Platform Approach |
The advent of high capacity field programmable gate arrays (FPGAs) with embedded memories, soft CPUs, DSP blocks, and programmable logic offer an alternative to the traditional design approach. These FPGAs can integrate into a single device all of the digital functionality traditional component collections provided (Figure 1). The CPU, the user interface controllers (i.e., LCD, touch panel), ASSP controllers for DSP functions and battery management, the glue logic, and the communications interfaces can all be implemented using library cores available from the FPGA vendor.
Cost-Effective Designs
The flexibility inherent in FPGA-based design also offers development teams the opportunity to create a single hardware platform that can serve as the basis for multiple products. For example, a hardware platform can contain an FPGA that implements all of the functions common to a portable medical device, such as display drivers, CPU, keypad interfaces, and generic digital input/output ports. Specific device designs can then augment this platform with whatever elements are unique to the device, such as analog input/output channels and DSP processing blocks. Reprogramming the FPGA handles required variations in specific functions such as display size, resolution, and the number of keys, which allows the hardware platform to remain unaltered.
This platform approach addresses the drawbacks that the traditional design approach faces. Component lifetime disappears as an issue: FPGAs typically have a market life in excess of 10 years. The traditionally arduous integration of components becomes a blending of library cores, which the FPGA design tools handle. A tool such as Altera’s SOPC Builder (system on a programmable chip) automatically stitch together the building blocks, including the developer’s own intellectual property, and configure the FPGA interconnects. Integrating all of the functions into the FPGA also addresses the power issue; a single FPGA uses less power than the component collection it replaces, and provides an opportunity to implement dynamic power management that further reduces system power.The platform approach can also reduce the high cost of certification. Once the platform has been certified, its use in subsequent designs requires only minor effort to obtain recertification. Similarly, upgrades and bug fixes can typically be accomplished by reprogramming the FPGA, even in the field. Such changes require far less testing to obtain recertification than hardware changes impose.
 Figure 2: An Altera FPGA-based reference platform |
The ideal place to start when adopting this platform approach is to use a reference design for portable systems from the FPGA vendor, as the initial platform (Figure 2). Reference designs accelerate the time to market for new designs because they have already been validated and include full board fabrication details. In addition, reference designs typically offer a library of fully-integrated FPGA cores that are tuned to the needs of portable systems.
Development teams wishing to make this shift to a platform design yet retain their software investments in specific processor technologies can adopt an intermediate approach by retaining the CPU that they are accustomed to and implementing all remaining circuitry in an FPGA. Ultimately, however, integrating all of the design’s logic functions into an FPGA provides the greatest benefit. By fully leveraging the capacity and flexibility of modern FPGAs, medical device designers can address the new requirements for device portability, intelligence, and connectivity that are challenging traditional design methods.
Online
For additional information on the technologies and products discussed in this article, see Medical Design Technology online at www.mdtmag.com or Altera Corp. at www.altera.com.
Charlie Jenkins is the senior technical product manager for Altera Corp., 101 Innovation Dr., San Jose, CA 95134. He is responsible for technical applications of programmable logic in medical & military uses. Jenkins can be reached at 831-419-4911 or chjenkins@altera.com.
