Select Page

Peter Marek, senior director embedded IoT, Advantech, explains the opportunities for AI to help quicken innovation

Time is of the essence

Such a bright and hopeful future for the medical industry is closer than we might think because AI is already reaping benefits for the sector around the world. Numerous players have established an ecosystem for medical AI and GPU and AI heavyweight NVIDIA is among the front-runners.

With the support of NVIDIA’s Inception project, a South Korean medical imaging start-up company called Lunit was able to create an AI-based system capable of detecting pneumonia in COVID-19 patients within seconds. The company said the beauty of AI in this application was its ability to process vast amounts of data and discover patterns, which would not be feasible otherwise, in terms of time required as well as skilled resources needed.

Another Inception member, the Australia-based start-up EMVision, was able to develop a lightweight brain-scanning device using the NVIDIA Jetson edge AI platform and NVIDIA DGX systems. The portable scanner – suitable for use at the point-of-care and in ambulances – is able to diagnose a brain stroke within minutes, which is massively important since the average stroke patient will lose nearly two million neurons for every minute the condition is left untreated. Time is clearly of the essence in such situations. Based on electromagnetic imaging technology, the helmet-like device of EMVision uses NVIDIA-powered AI to distinguish between ischaemic and haemorrhagic strokes (clots and bleeds) within the shortest possible timeframes.

Embedding inference

As the EMVision example shows, integration of AI technologies is making it possible to deliver better-than-ever patient care. The challenge for medical device manufacturers is to embed complex AI models that were tested and fine-tuned on server farms in data centres into comparably compact pieces of equipment. NVIDIA is supporting this with its Jetson platform optimised for embedded AI.

Advantech, as one of NVIDIA’s key ecosystem partners, is developing a range of off-the-shelf as well as custom embedded computing solutions using the Jetson architecture.

For example, based on NVIDIA’s latest Jetson AGX Orin, the AIR-030 high-performance edge AI box computers are capable of delivering powerful AI inferencing for energy-efficient smart machines. This system includes various I/O options, such as CAN Bus for motion and other control, LAN with optional PoE for camera input, and DI/O for sensor/actuators. Ideally suited for real-time video analytics and AI robotics applications, the AIR-030 series enables AI developers to boost performance by using 10GE for high-speed networking. Finally, the AIR-030 comes with an NVIDIA JetPack 5.0 software development kit (SDK) to minimise efforts for developers. Advantech’s DeviceOn for Edge AI package provides deployment, remote management and update capabilities and on NVIDIA platforms, this makes it possible to update AI inference models in bare metal as well as Docker environments. Powerful insights can be gained from remote monitoring of AI related performance indicators such as throughput levels.

Customised AI systems

Advantech’s Design and Manufacturing Services (DMS) have developed tailor-made AI solutions for use in the medical sector, from customisation of existing standard products to custom-designed boards and systems. With dedicated design resources and extensive experience in computer system design for medical environments, Advantech DMS builds on a solid foundation to take advantage of the latest AI technologies. Whether customers need inference solutions with efficient Jetson platforms, integration with FPGA devices or scalable high-end workstations that combine server processors with one or multiple GPU accelerators, Advantech’s engineers know how to create, verify, validate and certify these platforms to medical standards. The process includes design for manufacturing and for quality from the concept stage, with the active involvement of design engineers in terms of production and life cycle management.

To demonstrate the feasibility of AI inference in the healthcare industry, Advantech’s DMS team has developed a medical grade edge AI box PC based on NVIDIA technology. By integrating up to 384 high-power NVIDIA GPU cores with a high-performance CPU and a large memory subsystem, the PC delivers significant processing capacity for a variety of workloads such as image processing and fast data processing in video endoscopy, computed tomography (CT scans) or fluorescence tomography, while ensuring excellent energy efficiency.

Embedded IoT platforms

As a leader in embedded computing solutions with over 550 employees in Europe and more than 8,500 employees worldwide, Advantech supplies a wide range of powerful and scalable embedded Internet of Things (IoT) platforms, including motherboards, single-board computers (SBCs), computer-on-modules (COMs), box and panel PCs, servers, tablets, open-frame displays, monitors and video solutions. With a strong focus on the medical sector, Advantech offers long-term availability of products, optimal total cost of ownership (TCO) and support from initial product conception to end-of-life (EOL). Advantech’s medical customers, including the world’s 30 leading medical OEM companies, benefit from product lifetimes of up to 14 years as well as extended warranties, worldwide technical support and logistics services.

With a proven track record as a partner for manufacturers of medical equipment, as well as strong and continual investment in leading edge technology and dedicated development resources for tailor-made devices, Advantech is enabling OEMs to create innovative, ground-breaking medical solutions based on NVIDIA’s AI technology with low risk, minimal cost and reduced time to market. With such a contribution to the medical AI ecosystem, Advantech’s mission of “enabling an intelligent, sustainable planet” is helping the healthcare sector to incorporate innovation from research into advanced products for diagnostic imaging, therapy, robotics and patient monitoring. In turn, this will help to bring the power of AI into new areas such as endoscopy, radiology and surgery in the smart hospitals of the future.