How Premio’s KCO 3000 Series rugged edge computers enhanced 3D tomosynthesis platform’s real-time analysis beyond standard operating procedure
The Challenge
- By relying on consumer-grade compute systems, this 3D tomosynthesis innovator’s platform was limited in its ability to reach its full potential
- The short component life cycles inherent to consumer tech proved to be an ongoing issue that caused erratic revenue patterns
- A significant compute hardware upgrade that could easily be swapped was necessary to power the system’s ever-improving machine learning and AI capabilities
- In order to streamline product selection, supply chain management, and localized manufacturing in the United States, it was imperative that the compute vendor be ISO-13485 certified for medical device manufacturing
The Solution
- With 30+ years of design and manufacturing of industrial-grade computing solutions, Premio's advanced, rugged edge computers are highly suitable for medical applications
- Premio recommended its industrial-grade KCO-3000 series computers, perfect for the customer’s machine learning and AI application capabilities
- As a COTS product, the Premio certification-ready KCO-3000 industrial computer is designed for easy installation and interoperability with existing system components – making customization unnecessary and speeding time to market
- Premio’s 150,000 sq ft state-of-the-art ISO 13485-certified manufacturing facility in Los Angeles would enable flexible low-volume, high-mix manufacturing capabilities for custom system integrations
The Benefit
- Premio’s customer-centric and collaborative approach keeps the OEM in the loop at all times
- Premio provides turn-key system integration services in its own ISO-13485 facility, alleviating the need for initiating certification and speeding time to market
- Premio’s long-standing relationships with tech leaders gives them access to embedded roadmaps that allow them to plan for maximum product life cycles
- This engagement has fulfilled customer requirements and equipped its platform with the ability to deliver higher quality medical imaging for better patient outcomes
According to the CDC, breast cancer is one of the most common cancers in women in the United States. The American Cancer Society estimates about 287,850 new cases of invasive breast cancer will be diagnosed in women in the U.S. in 2022. Fortunately, breast cancer-attributed deaths have been decreasing steadily since 1989 – a decline of about 43% through 2020. This is often attributed to earlier screenings, increased awareness, and better diagnostics and treatments.
The Company
Today, technology allows healthcare practitioners to focus on what they do best – diagnose, care for, and treat patients with the best tools and knowledge available. Machine vision combined with GPU-powered inference analysis provides the predictive insights that garner better results. With technological advances in machine learning and artificial intelligence (AI), the medical world now has the ability to analyze this data in real-time to predict, prevent, and treat diseases without delay.
One medical device OEM has taken these advances to heart and is challenging the current standards of breast cancer surgery, including the excessive number of re-excision procedures performed every day. During breast cancer surgery, it is customary for the surgeon to include a border of tissue surrounding and beyond the edges of the visible cancerous tumor to ensure all cancer is removed. This border tissue, or the ‘surgical margin,’ is then examined by a pathologist who determines whether cancer cells are present. Results are labeled as Negative, Positive, or Clear by measuring the distance between the outer edge of the perceived normal tissue (the surgical margin) and the outer edge of the tumor containing cancer cells. Treatment decisions are influenced by these results. By deploying specimen tomosynthesis that leverages accurate, real-time, intraoperative margin assessment, surgeons – and their patients – have a clear advantage in obtaining negative final margins during the initial biopsy procedure, reducing the incidence of re-excision.
Through ongoing research and development of its proprietary 3D tomosynthesis technology, this company is empowering surgeons in the operating theater with real-time visualization of surgical margins. When valuable data is aggregated from edge computing machines, it delivers clear and immediate visibility that helps surgeons lower re-excision rates and preserve healthy breast tissue to improve patient outcomes.
The Challenge
Medical imaging and diagnostic devices are a complex lot. With compute power considerations, regulatory requirements, and now supply chain issues (among other hurdles), it’s no easy feat to navigate the path forward in bringing innovative and potentially lifesaving products to market.
As a leader in digital breast tomosynthesis, this particular OEM had developed breakthrough 3D technology that ups the ante for intraoperative quality of care. Yet the company’s existing computing platform limited its ability to reach the full potential of its technology. Reliance on consumer-level systems was an issue; without communication focused on revision control, revenue patterns would remain inconsistent based on excessively frequent product updates and limited component availability. While existing compute systems served their purpose initially, it became apparent that they did not offer the power, longevity, and ruggedness to take the medical imaging platform to the next level.
To quickly kick efforts into high gear, the OEM needed to partner with a reputable and agile vendor that was able to provide a superior drop-in replacement. If that partner also came through with a slate of wish list capabilities – like Bill of Materials (BOM) management, manufacturing quality controls, lifecycle management, and supply chain management fulfillment planning – the OEM would see long term competitive advantages.
One major requirement, however, was ISO 13485 certification – a regulatory mandate set forth by the Medical Device Act (Act 737) and designed to facilitate consistent standards for quality management systems (QMS). Having already achieve this certification, the OEM needed its ideal partner to hold these same credentials. The end solution would require a powerful ‘brain’ to extrapolate, analyze, and deploy data for an exceedingly high level of surgical accuracy. The OEM set out to find a vendor that would check all the boxes, but soon discovered not all industrial computing companies certified to critical medical standards possessed the level of expertise and support they were seeking.
The Solution
Enter Premio, a global solutions provider specializing in ruggedized computer hardware that not only happened to have ISO 13485-certified facilities, it had capabilities its competitors were lacking. With its customer-centric business model, Premio assembled a team consisting of sales, operations, technical/engineering, and quality control personnel to assess the OEM’s situation. By taking the time to understand the limitations of its consumer-grade compute equipment, Premio set out to determine the best option and transition its path to implementation.
Premio industrial computers are embedded computing solutions that act as either sub-assembly building blocks or final OEM system configurations in a broad range of key enterprise and IoT applications. System integrators and platform builders can accelerate time-to-market with Premio’s certification-ready industrial computers for new product introduction (FCC, CE, and UL). Industrial computers offer an ideal path to upgrade legacy computing platforms that require re-certification of sub-assembly builds in an accredited lab.
With a full slate of powerful rugged edge computing options in its lineup, Premio recommended its industrial-grade KCO-3000 series computers for running the OEM’s machine learning and AI applications which are so critical to surgical success. The KCO Series industrial computers are commercial off-the-shelf (COTS) computing solutions that offer reliability, regulatory safety, and embedded longevity with Premio’s extended lifecycle support. The KCO-3000 Series supports a Premio mATX Coffee Lake motherboard, Intel™ 9th Gen Core processors, and Nvidia RTX 4000/A4000 graphics card. As a COTS product, the Premio KCO-3000 is designed for easy installation and interoperability with existing system components – making customization unnecessary and a rapid time to market.
Premio's line of robust, purpose-built industrial computers is designed to enable medical inference at the edge. With high-performance GPU support, these certified, industrial-grade computing solutions are ready to meet the real time processing demands for medical imaging and diagnostics. The Premio KCO-3000 also offers the built-in hardware-based security and data encryption required in today’s medical applications through TPM (Trusted Platform Module).
Premio’s state-of-the-art ISO 13485-certified manufacturing facility would enable the flexible low-volume, high-mix manufacturing capabilities necessary for custom system integrations. With a focus on product revision control and lifecycle management, the customer could see its EOL issues disappear alleviating a major pain point that was limiting growth.
The Benefit
In Premio, this medical device manufacturer found the partner they had hoped for and then some. A one-stop shop offering everything from expert computing technology counsel and skilled operational manufacturing collaboration to COTS products and even post sale technical support. By converting the OEM’s existing computing platform to the KCO-3000 and manufacturing the new systems in its North American ISO 13485-certified manufacturing facility, Premio provides turn-key system integration services that include computer engineering design, BOM management and product revision control, manufacturing quality controls, lifecycle management, and SCM fulfillment planning.
Premio’s close and long-standing relationships with tech leaders means it is able to capitalize on embedded technology roadmaps for the long-term viability so critical to the medical market. These tight-knit relationships also mean that when something goes awry, Premio can readily rectify the situation. As part of the NVIDIA Partner Network that provides in-depth knowledge into the latest technologies powering machine learning and AI, Premio’s rugged edge computing solutions are validated to support enterprise performance acceleration cards (GPUs) from NVIDIA's diverse portfolio. For example, when a low end GPU used by this OEM went end-of-life (EOL) sooner than expected, Premio quickly advised the customer of its pending EOL status and overnighted a suggested replacement. They even secured inventory of the EOL GPU allowing time to validate and smoothly transition to the replacement GPU. With the previous vendor, the medical device company would have found out after the fact.
“From the very first meeting, Premio demonstrated a customer-centric approach to solving our problem. They listened intently to what we wanted to achieve, really taking the time to understand the issues we were facing from a compute standpoint,” said the medical device company’s R&D Director. “Throughout the entire process, Premio’s responsiveness and accessibility to personnel has been top-notch. They’ve kept us informed every step of the way and delivered as promised. With these assurances and expert support, we are able to advance performance, ensure product quality, and steadily increase revenue.”
In summary, Premio’s KCO-3000, high-speed computing and parallelized graphics computing have helped this OEM deliver a higher quality medical imaging platform for patient care that consumer-level computing products just can’t match.