Architectures
Architectures
Medical electronic devices cover a spectrum of equipment implementations, ranging from large diagnostic imaging machines that fill a room, to small mobile devices that patients carry with them.
Driven by healthcare dynamics for superior patient care at lower costs, providers are demanding increased visualization and transmission of medical image and video data from OEMs. Therefore, several technical trends are emerging that influence future equipment architectures:
- Movement towards commercial off-the-shelf (COTS) homogenous hardware platforms
- Migration from proprietary parallel to standardized serial component interconnects (HSIO)
- Increased wired and wireless device connectivity to the Internet
These trends, along with the spectrum of equipment implementations, are driving medical equipment toward one of three implementation architectures:
- Chassis Based -a passive backplane with heterogeneous or homogenous processing cards
- Modular Based - single board computer with PCIe* modules for application-specific functionality
- Portable Devices -FPGAs with embedded CPU(s), implementing all processing functions
Chassis
Large imaging systems have used chassis implementations to provide flexibility/scalability, offering hospitals/clinics multiple options at various price points. These systems are also used to accelerate complex hardware and software development across multiple platforms, teams, and locations.
Diagnostic imaging requires many electronic subsystems and specialized components to acquire and recreate the spectacular life saving pictures of the human body. These subsystems must meet the following requirements:
- Accommodate multiple cards for data acquisition, processing, control, and switch interconnects
- Transfer large amounts of data through proprietary parallel bus structures
Now you can implement commercial off-the-shelf (COTS) or proprietary chassis backplane architectures (see Figure 1) using standardized serial implementations with integrated serial transceivers [for example, PCI Express* (PCIe*) on Intel Arria and Intel Stratix FPGA families, which allow:
- Significantly increased subsystem data transfer rates
- Lower development costs and a smaller equipment footprint
- Higher performance, homogenous blade servers using CPUs coupled with FPGA coprocessing cards
Homogeneous processing architectures provide a scalable, ready-to-use hardware platform including user interface, storage, and connectivity with the capability to add algorithm acceleration through FPGA add-on cards. The OEM adds application-specific analog interface cards to complete the electronics of the system. See below.
Chassis-Based Architecture
Modular
Designers of medical modalities like ultrasound have historically implemented custom designed hardware. Now commercial off-the-shelf (COTS) modular systems are increasingly available based on single board computers (SBC) and embedded CPU boards, comprising either:
- x86 CPU-based technology with Windows OS
or - An open standard operating system, such as Linux*, coupled with RISC CPUs
These SBCs previously used PCI parallel standards as their I/O expansion interface for data movement, with much of the equipment functionality implemented in software. In the future, these SBCs will implement:
- Higher bandwidth PCI Express (PCIe) I/O, while maintaining PCI software compatibility
- Gigabit Ethernet for connectivity to other equipment and the Internet
For your application-specific needs, you can now:
- Easily implement PCI Express modular boards using the Intel® FPGA IP for PCI Express
- Integrate multiple serializer/deserializer (SERDES) channels over 10 Gbps within Intel® Arria® FPGA and Intel® Stratix® FPGA families
- Rapidly create custom functionality and interfaces using Intel Stratix FPGAs and the Intel® Quartus® Prime design software
Figure 1 shows how these modular PCIe function boards can be made application independent, providing higher volumes across multiple medical modalities by separating the application-specific analog interfaces to detectors/sensors onto daughter cards.
Modular-Based Systems
Portable
Portable handheld-size devices can implement mobile monitoring of patient vital signs, such as heart pacing, blood pressure, and glucose tracking.
These portable devices have been implemented using a host of discrete components and a CPU. It takes more time and resources to implement these custom hardware platforms, resulting in:
- Wasteful efforts developing standard blocks instead of custom application blocks
- Obsolescence risk from multiple discrete ASSP processor components
- Integration risk, cost, and inflexibility of ASICs
With the latest Intel Quartus Prime Design Software, Platform Designer, Nios® II embedded processor, SoC device family, and programmable hardware capabilities of Intel Cyclone®, Intel Arria® and Intel Stratix® FPGAs, you can:
- Integrate multiple ASSP devices and the processor into a single application-independent FPGA
- Implement independent functions into multiple Nios II processors on a single FPGA
- Generate coprocessing logic to “supercharge” functional performance
- Easily integrate all the application-specific and standard intellectual property (IP) blocks with Platform Designer
Additional productivity benefits include:
- Reusing previous internal engineering IP development efforts
- Leveraging additional Intel FPGA and partner IP blocks
- Separating application-specific analog I/O and sensor circuits onto modular daughter cards
The figure below shows how an Intel Cyclone or Intel Stratix series FPGA implements a system-on-a-programmable-chip solution using one of more Nios II processors, including custom logic. The user interface, application-specific analog functionality, and network connectivity can be added to complete the portable device.
Portable Devices
Related links
Medical Solution Reference Links
Additional Resources
Intel® FPGA Industry Applications
Learn how to leverage these application solutions to help meet your design challenges.
Intel® FPGA and Programmable Devices
Learn how these powerful devices can be customized to accelerate key workloads and enable design engineers to adapt to emerging standards or changing requirements.