Two recent trends in medical-imaging systems have created new cost and performance design issues for OEMs. First, a demand for high resolutions has led to the fusion of diagnostic modes and algorithm advances. The second trend is toward portable monitoring devices that require greater device intelligence and communications than previous equipment.

Field-programmable gate arrays (FPGAs) are flexible enough to meet the challenges of both trends. An FPGA is essentially an integrated circuit that can be programmed in the field after manufacturing. FPGAs offer high-value programmable technology, and enable near-zero design compromise. They are also obsolescence-free because they can be reprogrammed, letting designers continually update features and algorithms over a system's lifetime.

This flexibility lets designers create multiple versions of the same product or system for different applications when it is introduced. FPGA-based designs also let designers add new features and upgrades in response to changing market demands and standards with minimal engineering effort. This includes upgrading existing products in the field. By using device programmability, designers can continuously and cost-effectively refresh product lines and provide differentiating capabilities.

Getting the highest resolution

FPGA flexibility plays an important role in resolving design challenges as the industry strives to lower patient costs and detect diseases earlier using non-invasive methods. These new systems must process large amounts of data quickly and accurately. The electronics for such tasks use fine geometry micro-array detectors coupled with sophisticated software and FPGA-based hardware systems to analyze photonic and electronic signals. This includes scalable CPU platforms with FPGA support for data acquisition and co-processing. OEMs should consider certain factors to efficiently develop flexible, scalable medical-imaging designs. These include developing imaging algorithms, modality fusion, and scalable platforms. Developing imaging algorithms requires high-level intuitive modeling tools for continual improvements in digital signal processing (DSP). These advanced algorithms require scalable system platforms with significant increases in image processing performance, implemented in smaller, more cost-effective portable equipment.

Near real-time analysis requires system platforms that scale with both software programmable logic and and hardware CPUs. These processing platforms must meet various performance price points and be capable of bridging multiple forms of imaging information. FPGAs are easily integrated into multi-core CPU platforms, providing the DSP horsepower for the most flexible, highest performance systems. And to speed deployment and improve profitability, system designers must quickly partition and debug algorithms onto these platforms, using high-level development tools and intellectual property libraries.

Portability

Medical-system designers tied to inflexible traditional designs can benefit by shifting to the wider design latitude and cost efficiencies of FPGAs. Consider that a single FPGA device includes the tools needed for practically any system, including portable ones. The CPU, user-interface controllers, application-specific standard product controllers for DSP functions and battery management, glue logic, and communications interfaces are all implemented using library cores available from FPGA vendors and partners.

The design of a portable consumer monitoring device must include an efficient, rechargeable battery system, wireless communications, and built-in intelligence. Without careful control of the charging circuit's voltage and current, batteries can overheat and possibly explode. Discrete circuits used for active battery management present a further design challenge because they can be more costly than the battery pack.

Traditional design involves assembling a collection of CPUs, battery management, and interface devices, then connecting them with programmable glue logic. The other approach is a custom IC. This is impractical for cost-effective medical devices because of the relatively low production volumes and high development costs. Also, a traditional design incurs the high cost of regulatory certification. Each “point solution” design must undergo full certification, and component obsolescence forces periodic re-certification.

An FPGA platform resolves these traditional custom IC design disadvantages. Component lifetime becomes a non-issue because FPGAs have a market life in excess of ten years. Troublesome integration of individual ICs in a traditional design is replaced by a blending of library cores, which FPGA design tools handle efficiently.

These tools assemble the necessary building blocks, including the OEM's intellectual property. The tools then automatically configure the FPGA interconnects. Another plus: a single FPGA uses less power than the component collection it replaces and lets designers implement dynamic power management schemes, further reducing system power.

Further still, high certification costs are reduced. After certifying an FPGA platform, its use in subsequent designs requires only minor effort to recertify. Upgrades and bug fixes are handled by reprogramming the FPGA. Such changes require less testing to obtain recertification than hardware changes impose.