Medical imaging devices—such as magnetic resonance imaging (MRI), computed tomography (CT), and ultrasound systems—form an essential part of patient care in hospitals and clinics alike. Depending on their target use case, however, these systems require very different performance, robustness, and portability characteristics for integrated edge computers. By enabling compute-intensive processing directly on the device, medical imaging systems can also operate reliably without network availability, improving data security for compliance with strict medical privacy requirements.
In theory, fully custom monolithic architectures ensure maximum resource alignment for a given application and device form factor. However, this approach hinders parts replacement and design refresh that supports sustainability, cost efficiency, and right-to-repair initiatives that extend product lifecycles. Furthermore, changes to monolithic systems often require costly, time-consuming recertification that increases downtime.
To address this issue, equipment designers can turn to commercial-off-the-shelf (COTS) edge computing solutions. For example, open-standard computer-on-modules (COMs)—also known as system-on-modules (SOMs)—allow system architects to house all edge computing resources on a single board that is mounted on an application-specific carrier system. This approach promotes cost-effective modular repairs and performance upgrades to extend lifecycle, with open COM standards further easing sourcing efforts by reducing dependency on a single vendor.
For medical imaging systems, the core architectural decision lies in defining how computing resources, artificial intelligence (AI) functionality, and lifecycle management will be sustained over long operational lifetimes. This blog provides strategic recommendations on COTS edge computing solutions for different classes of medical imaging devices before briefly discussing networking and software considerations as medical settings embrace AI.
Handhelds – Extreme Portability through SMARC
Portable medical imaging devices include mobile ultrasound and Raster Scan Optoacoustic Mesoscopy (RSOM) systems that enable fast point-of-care diagnostics directly at the patient’s bedside or in emergency settings. These units combine high image quality with a compact form factor, but architectural decisions are not simply a matter of computing resource miniaturization. Here, system designers must determine how much AI inference, preprocessing, and connectivity is required within strict thermal and energy constraints.
The Smart Mobility Architecture (SMARC) COM open standard was specifically developed for compact, mobile embedded systems, making it particularly well suited for medical handhelds. With a credit-card-sized footprint of 82 x 50 mm, SMARC modules integrate easily into space-constrained enclosures. At the same time, the standard also provides extensive visual and display interfaces for high-resolution cameras and touch displays. The low power consumption of SMARC modules supports efficient battery operation, ensuring full mobility without compromising computing performance.
A solution that is ideally suited for these applications is the SECO SOM-SMARC-ASL. Based on Intel® Atom® x7000RE processors, it integrates Intel® UHD Graphics and offers a wide range of high-speed interfaces such as USB, PCIe, Ethernet, and MIPI-CSI, making it ideal for powerful edge systems. The module supports up to 16 GB of LPDDR5 memory and up to three independent 4K displays. The Intel® Atom processor enables computer vision and AI inference directly at the edge, allowing raw image data to be processed locally rather than transmitted to the cloud, enabling fast, privacy-compliant real-time diagnostics. Intel® Time Sensitive Networking (TSN) and Time Coordinated Computing (TCC) also ensure deterministic low-latency image data transfer and processing—critical for safe live imaging during ultrasound-guided biopsy, for example.
Mobile Carts – Boost Performance with COM Express and COM-HPC
Mobile imaging devices, such as bedside X-ray systems and increasingly portable MRI units, enable rapid diagnosis in intensive care or isolation wards without moving the patient, improving workflows and enhancing safety. For mobile carts, system architects should consider a forward-thinking approach to computing resource and I/O requirements: How can these scale to support new AI models, higher-resolution sensors, or additional accelerators throughout the product lifecycle?
For these systems, open COM standards remain an excellent choice for high imaging resolution and high data throughput. While still compact, COM Express and COM-HPC deliver significantly higher system performance than SMARC. COM Express provides a proven, robust standard with a wide range of I/O options, while COM-HPC is specifically designed for very high computing and bandwidth requirements, with substantially more high-speed interfaces.
Since mobile carts are powered by mains electricity or high-capacity batteries, COM Express and COM-HPC solutions are far less constrained by power consumption, enabling them to support high-performance processors and high-speed interfaces. Nevertheless, the resulting thermal load requires advanced thermal management to ensure stable and safe operation during time-critical live imaging.
For mobile imaging systems, the SOM-COMe-CT6-ASL, a COM Express Type 6 module that is also based on the Intel® Atom® x7000RE chipset, offers similar computing, graphics, and edge AI capabilities to the SOM-SMARC-ASL. Compared to the SMARC module, however, the COM Express SOM-COMe-CT6-ASL provides significantly more high-bandwidth I/O options to support high-resolution image and sensor data streams. These include up to six PCIe Gen3 lanes instead of four, up to eight Hi-Speed USB interfaces instead of six—plus an optional three USB 5 Gbps and two USB 10 Gbps interfaces—and up to two SATA Gen3 channels instead of one. While SMARC enables more compact, efficient designs, COM Express modules feature a larger 95 x 95 mm footprint (Type 6 Compact) with a carrier interface that is optimized for I/O scalability and performance headroom.
Alternatively, COM-HPC modules like SECO’s SOM-COM-HPC-A-RPL take performance a step further by integrating 13th Gen Intel® Core™ processors, supporting higher calculation capabilities, DDR5-5200 memory, and extensive higher-speed interfaces that include PCIe Gen4, USB4, and 2.5 Gigabit Ethernet. This makes COM-HPC ideal for applications such as high-resolution 3D reconstruction in CT or MRI systems, multi-stream video processing at high frame rates, and AI-assisted real-time diagnostics using dedicated GPU or PCIe accelerators.
Telemedicine – SBCs Accelerate Design and Setup
In addition to handheld and mobile imaging devices, telemedicine is becoming increasingly important for enhancing accessibility and efficiency to improve patient outcomes. Architecturally, telemedicine platforms are less concerned with direct sensor interfacing, instead stressing ergonomics and security in user-friendly workstations. As such, system designers must prioritize UHD multi-display support, reliable peripheral connectivity (camera/audio/HID), and long-term maintainability, with standardized I/O and controlled software updates.
Single-board computers (SBCs) are well suited for these use cases, offering simplified procurement and system integration over COMs, since all core computing and interface components are packaged on a single board. As compact, ready-to-use systems, SBCs are ideal for accelerating development of space-constrained medical applications. With that said, a lack of standardization across solutions can limit flexibility for upgrades or replacement, so use of open SBC standards like Pico-ITX is essential for long-term maintenance.
The SBC-pITX-ASL is an example of an SBC that offers this advantage. Like the SOM-SMARC-ASL and SOM-COMe-CT6-ASL, it is based on the Intel® Atom™ Industrial RE processor family and provides the same core capabilities for edge AI, TSN/TCC, and graphics, while eliminating the need for carrier design. The Pico-ITX standard—with its defined form factor of 100 x 72 mm and standardized interfaces such as dual 2.5 GbE, HDMI, USB 10 Gbps, SATA, and M.2—enables simplified upgrades and interchangeability between SBC generations during equipment refresh. As a result, telemedicine system architects can significantly reduce development effort and integration complexity when compared to COM-based or monolithic design.
Fixed suites – Industrial Edge PCs Support Advanced Imaging
Large stationary imaging systems like MRI and CT suites deliver high-resolution images and require powerful computing systems for real-time processing with high operational reliability. Due to their sensitivity to electromagnetic interference and heat, computing resources should be physically separated and medically isolated from scanners, which simplifies maintenance, cooling, and regulatory compliance. Consequently, the architectural priority is to successfully decouple certified imaging hardware from rapidly evolving computing resources.
With higher data rates and larger thermal design power (TDP) budgets than SBCs, industrial edge PCs present an ideal computing platform for fixed imaging suites. These self-contained solutions simplify maintenance through standardized component replacement and allow platform upgrades without modifying or recertifying the imaging device itself. Modular expansion also enables targeted upgrades, such as adding dedicated hardware accelerators for AI inference or 3D reconstruction.
The Palladio 500 RPL is a modular, industrial PC based on 13th Gen Intel® Core™ processors, delivering significantly higher computing and graphics performance than basic embedded computing boards. Extensive expansion options include PCIe Gen4, multiple 2.5 GbE ports, two DisplayPort interfaces, as well as USB, M.2, and wireless options, enabling powerful networking and video connectivity opportunities. Measuring approximately 240 x 143 x 267 mm, the Palladio 500 RPL leaves ample room for suite expansion, making it well suited for demanding edge and imaging workloads.
Networking and software integration in the age of AI
While edge computing enables secure, low-latency local processing that protects sensitive data and enhances patient safety, connecting imaging results to broader electronic medical record (EMR) databases is critical for efficient workflow across distributed teams. Nevertheless, direct EMR connection poses cybersecurity and operational risks in the event of compromised devices or high-volume data transmission, which can slow performance across EMR systems. For this reason, network segmentation and intermediary data buffers—such as picture archiving and communication systems (PACS) or vendor-neutral archives (VNA)—are often used to mitigate cyberattacks when uploading essential imaging data. When designing the software architecture of any medical imaging device, these hurdles must be considered to ensure patient safety and data security regardless of connection status.
In regulated medical environments, SECO’s Yocto-based Clea OS framework supports traceability, security hardening, and controlled software lifecycle management in line with long-term compliance requirements. This helps ensure devices continue to operate safely and securely both at the individual level and when connected to wider medical ecosystems. With AI at the forefront of many industries, the modular Clea framework also grants flexibility across different hardware and software solutions when deploying state-of-the-art workloads to assist in imaging diagnostics. While currently impeded by strict regulatory barriers that limit a rapid influx of AI functionality, medical imaging systems benefit from a software approach that preempts a more widespread adoption of AI in this space.
Conclusion
Medical imaging spans a wide range of devices—from handheld systems to large stationary scanners—each with distinct requirements for computing performance, power supply, and form factors, making tailored, integrated computing architectures essential. SECO supports these requirements across a wide range of COTS edge computing platforms that include COMs, SBCs, and industrial edge PCs. Moreover, SECO’s ISO 13485:2016 certification ensures high quality and safety standards for the development and manufacturing of medical electronics. In addition, SECO Clea OS provides a forward-thinking software framework to support robust cybersecurity and AI integration for next-generation medical imaging devices.
Contact SECO to discover the right Intel-based edge computing solution for your medical imaging application.