Category: Health

The use of edge AI in medical imaging offers the possibility of enormous benefits to stakeholders throughout the healthcare sector.

On the provider side, edge AI imaging can improve diagnostic accuracy, boost physician efficiency, speed case processing timelines, and reduce the burden on overstretched medical personnel. Patients benefit from shorter wait times for their diagnostic test results and a better overall quality of care.

But it can be challenging to develop AI-powered solutions needed to make this promise a reality. The computing requirements to implement edge AI in medicine are high, which has historically made it both difficult and expensive to obtain adequate computing resources. It can also be hard to customize the underlying hardware components well enough to suit medical imaging use cases.

It’s a frustrating situation for anyone wanting to offer innovative AI-enabled imaging solutions to the medical sector—because while the market demand certainly exists, it’s not easy to build products that are effective, efficient, and profitable all at the same time.

But now independent software vendors (ISVs), original equipment manufacturers (OEMs), and system integrators (SIs) are better positioned to innovate edge AI-enabled medical imaging solutions. The prevalence of rich edge-capable hardware options and the increasing availability of flexible AI solution reference designs make this possible.

AI Bone Density Detection: A Case Study

The AI Reasoning Solution from HY Medical, a developer of computer vision medical imaging systems is a case in point. The company wanted to offer clinicians an AI-enabled tool to proactively screen for possible bone density problems in patients so that timely preventive steps could be taken.

They needed an edge AI deployment that would put the computational work of AI inferencing closer to the imaging devices, thereby reducing network latency and bandwidth usage while ensuring better patient data privacy and system security. But there were challenges.

The edge computing power requirements for a medical imaging application are high due to the complexity of the AI models, need for fast processing times, and sheer amount of visual data to be processed.

In addition, special challenges involved developing an AI solution for use in medical settings: an unusually high demand for stability, the need for waterproof and antimicrobial design elements, and the requirement that medical professionals approve the solution before use.

The solution can automatically measure and analyze a patient’s bone density and tissue composition based on the #CT scan data, making it a valuable screening tool for #physicians. HY Medical (Huiyihuiying) via @insightdottech

HY Medical leveraged Intel’s medical imaging AI reference design and Intel^® Arc^™ graphics cards to develop a solution that takes image data from CT scans and then processes it using computer vision algorithms. The solution can automatically measure and analyze a patient’s bone density and tissue composition based on the CT scan data, making it a valuable screening tool for physicians.

The solution also meets the stringent performance requirements of the medical sector. In testing, HY Medical found that their system had an average AI inference calculation time of under 10 seconds.

Intel processors offer a powerful platform for medical edge computing, which allows the company to meet its performance goals with ease. Intel technology also provides tremendous flexibility and stability, enabling the wide-scale application of this technology in bone density screening scenarios.

Reference Designs Speed AI Solution Development

HY Medical’s experience with developing their bone density screening solution is a promising story—and one that will likely become more common thanks to the availability of AI reference designs. These reference architectures make it possible for ISVs, OEMs, and SIs to develop medical imaging solutions for a hungry market both quickly and efficiently.

Intel’s edge AI inferencing reference design for medical imaging applications supports this goal in several ways:

Tight integration with high-performance edge hardware: Ensures that solutions built with the reference design will be optimized for computer vision workloads at the edge. The result is improved real-world performance, better AI model optimization for the underlying hardware, and increased energy efficiency.

Flexible approach to AI algorithms: Because different software developers work with different tools, multiple AI model frameworks are supported. Models written in PyTorch, TensorFlow, ONNX, PaddlePaddle, and other frameworks can all be used without sacrificing compatibility or performance.

AI inferencing optimization: The Intel^® OpenVINO^™ toolkit makes it possible to optimize edge AI models for faster and more efficient inferencing performance.

Customized hardware support: The reference design also factors in the special needs of the medical sector that require customized hardware configurations—for example, heat-dissipating architectures, low-noise hardware, and rich I/O ports to enable connection with other devices in clinical settings.

The result of reference architectures such as this one is that they shorten time-to-market and reduce the inherent risks of the product development phase, giving innovators a clear path to rapid, performant, and profitable solution development. That’s a win for everyone involved—from solutions developers and hospital administrators to frontline medical professionals and their patients.

The Future of AI in Medical Imaging

The ability to develop innovative, tailored solutions quickly and cost-effectively makes it likely that far more AI-enabled medical imaging solutions will emerge in the coming years. The potential impact is huge, because medical imaging covers a lot of territory—from routine screenings, preventive care, and diagnosis to support for physicians treating diseases or involved in medical research.

Hospitals will be able to use this technology to improve their medical image reading capabilities significantly while reducing the burden on doctors and other medical staff. The application of edge AI to medical imaging represents a major step forward for the digital transformation of healthcare.

Edited by Georganne Benesch, Editorial Director for insight.tech.

Healthcare professionals have a singular mission: provide the best possible care for patients. But from admittance to discharge and everything in between, they face countless challenges.

Persistent staff shortages, constrained resources, and tight budgets are just a few. The greatest challenge is access to essential information about a patient’s condition throughout their hospital journey, specifically the second-by-second time series waveform data generated from the biomedical devices monitoring a patient. When seconds matter, how do hospitals harness this data and make it easily accessible to their healthcare teams?

Why Time Series Data Matters

The answer to this challenge is a single, open platform that continuously collects, processes, and unifies disparate data and presents it to clinicians in real time. Take, for example, an eight-hospital system in Houston that was confronting staffing issues and limited provider coverage—especially overnight. This forced difficult decisions, like hiring more travel nurses and physicians or turning patients away. All that changed when the organization implemented the Sickbay^® Clinical platform, a vendor-neutral, software-based monitoring and analytics solution, from Medical Informatics Corp. (MIC).

The platform enables flexible care models and the #development and deployment of patient-centered #AI at scale on a single, interconnected architecture. @sickbayMIC via @insightdottech

Sickbay is an FDA-cleared, software-based clinical platform that can help hospitals standardize patient monitoring. The platform enables flexible care models and the development and deployment of patient-centered AI at scale on a single, interconnected architecture. Sickbay redefines the traditional approach of storage and access to static data contained in EMR systems and PACS imaging. The web-based architecture brings near real-time streaming and standardized retrospective data to care teams wherever they are to support a variety of workflows with the same integration. This includes embedded EMR reporting and monitoring data on PCs and mobile devices.

“Out of about 800,000 data points generated each hour for a single patient from bedside monitoring equipment, only about two dozen data points are available for clinical use,” says Craig Rusin, Chief Product & Innovation Officer and cofounder at MIC. It’s not widely known that alarms from non-networked devices such as ventilators outside of a patient’s room are difficult for staff to hear or view remotely. Similarly, current patient monitoring doesn’t use AI tools with the existing data to inform patient care.

Measuring the Impact

Hospitals and healthcare systems using Sickbay have redefined patient monitoring and have created a new standard of flexible, data-driven care by demonstrating the ability to:

• Rapidly add bed and staff capacity while creating flexible virtual care models that go beyond traditional tele-sitting, admit, and discharge.
• Provide more near real-time and retrospective data to staff already on unit, on service, or on call to improve their workflows and delivery of care.
• Create virtual nursing stations where one nurse can monitor 50+ patients on a single user interface across units and/or facilities.
• Leverage the same infrastructure to create virtual command centers that monitor patients across the continuum of care.

No matter the method of deployment, Sickbay gives control back to healthcare teams and provides direct benefit back to the hospital. Benefits reported include reduced labor, capital, and annual maintenance costs as well as improved staff, patient, and family satisfaction. Most important, clients using Sickbay see direct impact on improvements in quality of care and outcomes, including reductions in length of stay, code blue events, ICU transfers, time on vent, time for dual sign-off, and time to treat.

Results such as these provide the pathway for other hospitals to rethink patient monitoring and realize the vision of near real-time, patient-centered AI. Healthcare leaders have proven that going back to team-based nursing by adding virtual staff can help reverse the staffing crisis. “This isn’t about taking nurses away from patients. This is about taking some of the tasks and centralizing them,” says Rusin. “There will never be enough nurses, physicians, and respiratory therapists to cover all of the demand required for the foreseeable future. We need to get bedside teams back to bedside care. Flexible, virtual care support makes that a reality.”

Changing the Economics of Care

Sickbay provides the ability to change the economics of monitoring patients and directly impact improvement in quality and outcomes.

The ability to integrate with different devices, regardless of function or brand, is the key. “We have created an environment that allows our healers to get access to data they have never had before and build content on top of that, in an economically viable way that has never been achieved,” Rusin says.

For healthcare providers, having the data available is game-changing, says MIC EVP of Strategic Market Engagement, Heather Hitchcock. As one doctor commented: “In a single minute, I have to process 300 data points. No machine is ever going to make a decision for me, but Sickbay helps me process that data faster so I can make the right decision and save more lives.”

From Scalable Patient Monitoring to Predictive Analytics

Sickbay’s value extends beyond near real-time patient monitoring and virtual care to long-term treatment improvements. Sickbay supports the ability to leverage the same data to develop and deploy predictive analytics to help get ahead of deterioration and risk.

Clients currently and continuously develop analytics on Sickbay. For example, one client integrates 32 near real-time, multimodal risk scores into its virtual care workflow. Another client created a Sickbay algorithm that analyzes data generated by two separate monitoring devices to determine ideal blood pressure levels in patients. “The particular analytic requires the blood pressure waveform from a bedside monitor and a measure of cerebral blood density from a different monitor,” says Rusin.

Saving Lives with Data

Treatment of patients across the care continuum today will lead to improved care tomorrow. To do that, reliable, specific data is the very starting block. Without it, clinicians are left to their best guesses to solve the body’s most urgent care needs without the data-driven decision-making support they desire. That’s slow, costly, unfair to caregivers, and ultimately not providing the best benefit for the patient.

To truly realize a future where treatment is as specific and individual as the person it serves, healthcare must harness patient data in a way that is most impactful—specific, accurate, near real-time, vendor-agnostic, transformable, and instantly accessible. Leveraging the power of time series data empowers healthcare providers to help more people than it ever has before, and more effectively. After all, saving lives is healthcare’s primary mission.

Edited by Georganne Benesch, Editorial Director for insight.tech.

X-rays, ultrasounds, and endoscopy machines generate massive volumes of data—sometimes too much to make sense of. In response, medical device OEMs integrate AI directly into medical imaging and diagnostics machines to make screening procedures more efficient, effective, and accessible for clinicians and patients alike.

Supporting AI-enabled medical imaging and diagnostics requires high-end hardware with the graphics and compute performance to execute intelligent imaging workloads in real time. Until recently, the easiest way to enable these capabilities was through discrete accelerators—an approach that can be expensive and inefficient in terms of upfront hardware costs and power consumption.

But by far the most costly design decision is the wrong system architecture. AI is evolving rapidly, so without flexible, adaptable, and upgradable system hardware, equipment can become obsolete before it is adequately broken in.

“AI workloads are advancing so quickly, it’s sort of dangerous when you start talking about hardware at all,” says Zeljko Loncaric, Market Segment Manager for Infrastructure at congatec AG, a global leader in embedded solutions. “That’s one of the most significant challenges facing medical device designers. They also face hurdles in implementing newer functionality in long-lifecycle systems.”

COM Express modules based on Intel^® Core^™ Ultra mobile processors address these challenges. They offer superior performance and efficiency in AI workload processing thanks to integrated GPUs and NPUs. And their inherent modularity streamlines the initial design process while enabling easy upgrades, processor generation over processor generation.

#AI #technology represents a meaningful advancement for #medical imaging, with the potential to significantly improve diagnostic efficiency and accuracy. @congatecAG via @insightdottech

Balancing Edge AI Longevity and Innovation in Embedded Computing

Because medical imaging devices must undergo a comprehensive certification process before they can be used, their lifecycles tend to average a decade or more. Meanwhile, AI technology represents a meaningful advancement for medical imaging, with the potential to significantly improve diagnostic efficiency and accuracy in ultrasounds, mobile ultrasounds, endoscopy machines, X-rays, and more.

But faced with the time and expense of redesigning and recertifying a medical device, OEMs hesitate to transition to next-generation platforms that support AI without an extremely compelling business case. And without being able to answer how long a system design will remain relevant, that business case becomes less compelling.

Enter new Intel Core Ultra Mobile processors, the first x86 processors to integrate an NPU, and one of the most power-efficient SoC families on the market today. The integrated NPU enables support for advanced AI workloads without the added cost and complexity of a discrete accelerator. When paired with the SoC’s leading performance-per-watt, medical device designers can better manage power consumption and thermal efficiency in resource-constrained edge AI deployments.

“The processor’s per-watt performance is also highly interesting in the context of mobile ultrasound devices and other battery-powered systems,” notes Maximilian Gerstl, Product Line Manager at congatec. “What Intel did with the architecture is very impressive. The numbers look great in terms of performance—not only on the CPU side, but also in terms of graphics. The new processors also offer an unprecedented level of flexibility to customers, allowing them to upgrade their systems across multiple generations while retaining the same form factor.”

“If there’s not a great new technology coming up, organizations will stay on the same module for 10 years or more so that they don’t have to recertify,” he continues. “Intel Core Ultra Mobile processors are a big step up. Healthcare organizations will have to think about changing to it.”

Open-Standard Modules Fast-Track System Upgrades

The latest congatec conga-TC700 COM Express Compact module incorporates the processing performance and application-ready AI capabilities of Intel Core Ultra Mobile processors in a plug-and-play form factor. Medical device designers can leverage the module as a shortcut to building efficient edge AI systems while significantly improving time-to-market and reducing total cost of ownership (TCO). And since COM Express is an open hardware standard governed by the global technology consortium, PICMG, the TC700 provides a vendor-neutral path to system upgrade whereby a legacy module can simply be swapped out for a higher-performance one with the same interfaces.

“The ability to quickly swap hardware means an organization can have its applications running for a very long time,” Gerstl explains. “Though they have to recertify new hardware components, they can bring over a lot of their software and hardware designs from previous applications.”

Intelligent Healthcare, Enabled by Edge AI Solutions

The conga-TC700 is supported by congatec’s OEM solution-focused ecosystem, which features efficient active and passive thermal management solutions, long-term support, and ready-to-use evaluation carrier boards. The company is also exploring how the open-source Intel^® OpenVINO^™ toolkit can empower its customers in the development and deployment of AI vision systems. According to Gerstl, the company is working on early benchmarking with specific use cases to help customers get their applications up and running more quickly.

For congatec, the availability of Intel Core Ultra Mobile processors represents a considerable step forward in the price, performance, and power consumption of next-generation edge AI devices. For medical device OEMs, these processors provide a compelling path to new, AI-enabled imaging and diagnostics equipment.

“We will continue to enable AI acceleration, hardware, and software and bring it to our products,” Gerstl says. “We want to enable this new trend.”

This article was edited by Georganne Benesch, Editorial Director for insight.tech.

In the healthcare industry, there are companies that build medical equipment and providers that use these machines and devices. While the business models of these two groups are completely different, they are guided by the same challenges and opportunities. Both are eager to deploy the latest technologies but are faced with strict regulatory requirements and short product life cycles.

A traditional medical device is designed with data residing locally on the device. With the increasing demand of interoperability within healthcare facilities, healthcare professionals can improve efficiency effectively with AI-enabled medical PC solutions designed to sustain the mission-critical environment and process patient data throughout the treatment process.

Managing all these constraints is a tall order, but today’s medical computers are up to the task. Hygienic, compact, and portable, medical-grade AI computers can be used by practitioners throughout hospitals and clinics. And high-performance processors enable near-real-time AI analytics, helping doctors and nurses make faster, better-informed diagnostic and treatment decisions.

#EdgeAI and #ComputerVision have become increasingly important to today’s imaging and patient monitoring machines, which can swiftly analyze #data and support physicians with diagnoses. @OnyxHealthcare1 via @insightdottech

Keeping Up with AI Innovations in IoT Medical Devices

Edge AI and computer vision have become increasingly important to today’s imaging and patient monitoring machines, which can swiftly analyze data and support physicians with diagnoses. But for medical device development, incorporating these cutting-edge capabilities can be a struggle. Medical device development takes on average from eight to 24 months to implement hardware and software design changes in accordance with regulations, and another two to three years to obtain certification via clinical trial.

“They don’t have the luxury to continuously upgrade to the latest technology,” says John Chuang, President of Onyx Healthcare, Inc., a Taiwan-based global producer of medical PCs and hospital IT solutions.

And once those finished medical devices are released, they need to stay in service for a long time. Hospitals have a complex mix of technology, and don’t usually upgrade their equipment for 10 years or more—an eternity in the fast-moving world of medical AI and computer vision development.

To keep machines as up-to-date as possible, Onyx collaborates with medical device companies, hospitals, and Intel, which supplies the processors for the all-in-one (AIO) medical computer the company produces for hospitals and clinics. Intel high-performance processing power is the key that enables software to run edge AI analytics.

Working closely with Intel, Onyx can provide a scalable custom design that allows medical device companies to incorporate the latest processors into its medical-grade computing technology. “By providing the latest technology to medical OEMs and ODMs, we help them keep a step ahead, so they don’t have to worry their technology is outdated by the time their devices are launched,” Chuang says.

Delivering Machine Information Where It’s Needed

In hospitals, medical devices are part of an elaborate symphony that requires precise timing and coordination. Doctors rely on information from many sources to diagnose and treat patients, including medical records and lab results, blood pressure and oxygen monitors, and images from X-ray, CT, and ultrasound scanners. But since these machines are made by different manufacturers and use different software protocols, they typically don’t connect with one another—or with hospital IT systems. As a result, doctors often must examine disjointed patient data.

A system like the Onyx AIO medical AI computer serves as a symphony conductor, integrating data from all sources—including patient records and off-site machines. It enables the transmission of high-resolution images and the performance of AI analytics, giving doctors a comprehensive, near-real-time view of a patient’s condition.

“The data transmitted is informative enough for physicians to make sound, timely treatment decisions. That’s especially crucial for patients in critical care, and in situations where the doctor needs to determine whether surgery is required,” Chuang says.

The ONYX AIO AI computer is also designed to meet hospitals’ rigorous sanitary requirements. For example, instead of using a fan for cooling, it uses an onboard heat sink, creating a closed system that won’t transfer germ-carrying air into hospital corridors or patient rooms. “We are able to use a fanless design because of the efficiency of low-wattage Intel processor technology,” Chuang says.

Medical IoT in Action: Mobile Nursing and Telehealth Solutions

Connecting patient information via medical computers can help hospitals and clinics achieve greater interoperability. That’s an important goal for the CAIH, a French government alliance formed to consolidate technology requirements across the country’s hospital networks. Onyx developed two solutions to help the organization achieve its objectives.

The first is mobile nursing stations—carts containing an AIO AI computer that nurses can bring on their rounds. The medical computer enables them to keep an eye on every patient under their care as they go from room to room. In addition to keeping nurses apprised of patients’ vital signs, the AIO helps monitor equipment, letting nurses know, for example, if an IV is running low on fluid.

AI monitoring helps short-staffed hospitals better attend to patients’ needs, Chuang says. It also helps them deal with the fast-growing use of telehealth. In a second solution it developed for the CAIH, Onyx enables AIO computers to connect doctors with patients, caregivers, and medical equipment at remote facilities—including skilled-nursing homes, where a physician may not be present.

Doctors can view patients from their own AIO computer and guide nurses in using medical instruments, such as portable ultrasound machines or scopes for examining the ear, nose, throat, or skin. Devices are equipped with high-definition cameras that relay medical-grade images to the doctor.

“With this information, physicians can do some diagnostics and quickly determine whether a patient needs to come to the hospital right away,” Chuang says. Otherwise, many would have no choice but to be transported there—often a challenge for those in a skilled-care institution.

Onyx AIO computers are also enabled for 5G communications, allowing remote facilities with a 5G network to relay alerts for patient vital data or slip-and-fall accidents directly to doctors or nurses, instead of waiting for the information to be processed in the cloud.

Building Future-Ready Technology

As AI capabilities expand, medical computers are assuming a greater role in patient care. But to stay useful, they must evolve along with the machines they connect with, Chuang says.

“Medical computers need to become more like medical devices themselves. We’re seeing greater demand for them to interface with specialized machines, and demand for data processing speed is also increasing. By building the latest Intel technology into our computers, we are able to satisfy those needs,” Chuang adds.

This article was edited by Georganne Benesch, Editorial Director for insight.tech.

Breakthroughs in biomedical research often come from understanding correlations and causality—the what and how of the body’s physiological processes.

Scientists make observations such as a higher rate of cancer or better reaction to vaccines by correlating data sets. They then research underlying reasons for these correlations. Systematically plodding through these cycles of biomedical research is tedious but rewarding work.

Now AI-driven computer vision applied to medical imaging accelerates discovery of data correlations. It finds problem points worth exploring much more quickly. In doing so, AI helps humans zero in on problems faster. And that can help human scientists arrive at life-changing medical solutions much more quickly.

AI Models in Medical Imaging

One use case is how Mayo Clinic uses AI and machine learning to profoundly enhance the capabilities of ultrasound imaging. As a starting point, the medical institution uses the latest technologies, tools, and products from Intel and Dell—the Intel^® Geti^™ platform and the Intel^® OpenVINO^™ toolkit running on Dell edge systems—to find kidney stones from endoscopy videos of the organ, and to assess Point of Care (POC) Ultrasound images for cardiac function.

Mayo Clinic’s work in the use of AI ultrasound imaging is a particularly useful case of the technology, says Alex Long, Global Head of Life Sciences Strategy at Dell Technologies, a solutions provider that offers an extensive product portfolio and comprehensive services. For too long, interpretation of ultrasounds has been subjective, prone to errors, and requires specialized training.

Visual #AI models, trained on banks of #data, can help providers offer more personalized care at the bedside. Augmenting care with AI can find anomalies faster, more accurately, and with minimal training. @DellTech via @insightdottech

But visual AI models, trained on banks of data, can help providers offer more personalized care at the bedside. Augmenting care with AI can find anomalies faster, more accurately, and with minimal training. Modern approaches which leverage pre-trained models and active learning enable the rapid development and deployment of these models. “Our care providers understand the benefit of using AI to aid in patient care, but in cases like the POC ultrasound, there wasn’t a viable AI model available,” says Dr. David Holmes of Mayo Clinic. His team of engineers leveraged interactive AI modeling tools to rapidly develop an AI solution that assess the quality of the images at the bedside in order to ensure the best images are used in the patient care.

The use of AI in medical imaging is about more than its capacity as a diagnostic tool. “It’s about leveraging visual AI to interpret imaging data and to accurately augment the capabilities of the human,” Long says. Diagnosticians trained to sift through files to find problems—evaluating mammograms to find early signs of breast cancer is a good example—can also benefit from AI guiding them to more places to evaluate. The advantage of AI is that it finds patterns that the human eye, due to confirmation bias, might miss.

A variety of additional scenarios in biomedical research can benefit from AI, especially if they involve imaging data. “It turns out there’s a lot of other medical systems that are visual in nature,” Long says. And they could all benefit from using AI as a tool to augment human abilities.

Collaboration Propels Innovation

A partnership between Intel and Dell Technologies enables these AI-driven breakthroughs. “The definition of community is a group of people with like-minded aspirations who are trying to achieve a goal together,” Long says. “We’re seeing a healthcare life sciences community being born between Dell and Intel.”

Collaboration between the two companies has evolved organically over many years, and Dr. Holmes’ work is one example of how the two bring their strengths to the table. The companies’ healthcare solution teams and their technology and product platforms enable collaborations with leading biomedical researchers and providers.

“The depth of our portfolio, the depth of our partnership, and the expertise in IT and infrastructure required to deliver” are what Dell brings to the table, Long says. In addition, Dell keeps in mind that the healthcare industry places heavy emphasis on privacy and protection of sensitive patient health information. “It’s not just about technology adoption to mitigate costs,” Long says, “it’s about technology to advance the human initiative of improved health. We’re passionate about really advancing the care of human beings.”

The Future of AI in Healthcare

The Mayo Clinic use case offers a glimpse of what is possible with AI models in biomedical research. We are just beginning to explore ways that AI can find correlations on visual imaging data, directing humans to new avenues for further exploration.

Researchers almost always try to find correlative data to drive conclusions, and “if you want something to identify a correlation, there’s nothing better than AI,” Long says. “I’m very excited about AI’s potential in accelerating diagnostics, improving patient care, and rapidly getting to understand the next wave of heuristics and treatments.”

When it comes to the human body, there’s a lot left to discover. It’s an exciting time to work at the intersection of technology and medicine because the volume of discoveries that AI can facilitate is simply mind-boggling. AI can train its eyes on years of data. The results are likely to be nothing short of revolutionary.

Edited by Georganne Benesch, Editorial Director for insight.tech.

When you or a loved one faces surgery, you naturally want to make sure that you have the most skilled surgeon. Not to mention that it is in the most up-to-date facility with the most sophisticated technology to ensure the best possible outcome. After all, the stakes can be incredibly high. So it’s disconcerting to think that, until recently, surgeons lacked decision-support resources in the OR.

While medical use of technology has evolved rapidly over the past few years, the surgical field has been a little slower to adopt these advances than some other industries. Surgeons are used to relying on the skill of their hands and the knowledge gained through experience—with good reason. But medical technology isn’t all about robot arms and AI-guided surgery; there’s a lot to be gained just by freeing healthcare data from its traditional silos and giving surgeons access to that information where and when they need it most—in the OR.

We talk more about this with Dennis Kogan, Founder and CEO of digital surgery platform provider Caresyntax, as well as the dynamic challenges of the operating room, the importance of a good partner ecosystem, and how AI-assisted surgery can improve patient outcomes (Video 1).

Video 1. Dennis Kogan, CEO of Caresyntax, discusses the integration of real-time AI-driven data into surgical procedures, emphasizing its critical timing and impact on surgeons. (Source: insight.tech)

How are technological advancements in the OR changing healthcare expectations?

My dad is a surgeon, and years ago when I was in college, I was talking to him about how much decision support athletes get around things like performance management, situational awareness, and analytics. And he told me, “We have nothing like this in surgery. We have very interesting and important medical devices, and we’re continuously getting clinical innovation into our hands, but there isn’t really a lot of data-usage and decision-making support.”

And up until a few years ago that hadn’t changed much. There was a ton of innovation around medical devices, but at the end of the day that was still helping only how surgeons operated with their hands. The advancements that we see now enable surgical teams to have better decision-support mechanisms as well.

I think there is more and more expectation that surgeons cannot just be thinking about the risks of the procedure by themselves. And they do want support; they do want additional information to stratify risks more. And doing it all in their heads is probably no longer acceptable anymore.

What are some of the challenges integrating new technologies into the OR ?

Relative to other types of therapies, patients are probably less aware of what’s happening in the OR. Naturally—they’re under anesthesia. What they want is to understand how likely they are to have a good outcome. And I think they would probably be surprised that not as much integrated decision-making support is available to their surgical teams as they would expect.

The challenge to innovating the surgical field with technology is that surgery is a real-time intervention, and you have to integrate the AI and the software so that it runs in that setting. There should be almost no lag time in the OR. And that by itself is a higher hurdle than for a lot of other information technology used in healthcare. Of course, there is also a pretty high threshold for quality and operational effectiveness.

The surgical environment is also extremely dynamic. So how does a surgeon adapt to a changing clinical picture during the procedure? And it’s not only quantifiable activities and techniques; there’s also communication and teamwork. Surgery is actually a team sport. Part of the outcome depends on how well a surgeon does a certain maneuver, but another part of it is how well they communicate with the nursing staff and anesthesiologist. It’s so complex that it’s almost impossible to foresee how it could be replaced by artificial intelligence in the foreseeable future.

But AI does have a lot to give in terms of bringing the right information and options to the fingertips of physicians, just because of that dynamism. In one day a surgical team may be operating on very different types of patients: a healthy 25-year-old female and then a very sick 85-year-old male. The team has to be able to adjust a lot of inputs and make a lot of decisions.

That cognitive overload can cause suboptimal decisions or mistakes. Probably one out of seven cases has some sort of significant complication—over 15%. And so what we’re talking about here is proactive risk management through situational awareness—through automation. It’s about reducing and removing unwarranted variability driven by cognitive overload and a changing clinical picture. The best use cases we see right now for AI are in showcasing specific information about a given patient and a given procedure to be able to guide the entire pathway for that procedure and have the outcome be better than it would have been without that support.

“The challenge to innovating the surgical field with #technology is that surgery is a real-time intervention, and you have to integrate the #AI and the #software so that it runs in that setting” – Dennis Kogan, @caresyntax via @insightdottech

What is the benefit of combining AI with patient data?

First and foremost, truly integrated surgical-decision support touches on all points of the peri-operative cycle. Because everything that happens before and after a surgery is also extremely important, the best-integrated platforms allow for connectivity between the operating room and the pre- and postoperative spaces, times, and activities.

There are decisions made right before the patient enters the operating room—preparing the right tools, the right medications, having the right people at the table. It also includes the electronic medical record, because that has a trove of data about the patient and his or her predispositions. Then there’s the situation inside the OR, where medical devices and video cameras can be connected. And then afterward: knowing what level of risk that patient is exiting the OR with may change the protocol of how they are going to be taken care of. Maybe they can go home; maybe they need to be in the ICU; maybe they need an extra dose of antibiotics.

So to get the best, smartest insights you have to have a full peri-operative clinical and operational record, but the crown jewel is the intra-operative space—because that of course is the most mission-critical piece, where things can really go wrong. And because of that, and because the OR is real time, it requires an additional level of sophistication. And it’s not, in technical terms, a cloud-friendly territory. It’s all on the edge, because you cannot rely on two-second upload and download from a cloud. So edge computing and the Internet of Things technology toolkit are extremely important here.

At the same time, this technology solution has to be very robust and attractive from the perspective of deployment and cost. Because at the end of the day, anything that is overly expensive or unwieldy—another huge machine being rolled into an already very packed operating room—is just not going to work.

It took us at Caresyntax—with the help of a few technology partners—years to develop this platform in a way that achieves all these parameters. But I do know that it’s possible. Things are still sort of at the beginning, but I think the next decade will probably see every OR being equipped with these kinds of systems. And in 10 years physicians will be wondering how they were working without it.

How can hospitals future-proof this kind of investment?

Every industry goes through a cycle of having a few vendors create kind of a walled garden at first, and then gradually users expect more and more flexibility to add value and to add new applications. I think surgery and healthcare will need to undergo the same change.

The medical-device world has a lot of proprietary intellectual property, for some good reasons. Historically that’s been a dominant mindset for physicians, too—thinking of the operating room through the prism of a device and a vendor, to a certain degree. So the first investment that needs to be made is in reinventing and recalibrating that mindset. The operating room should be seen not as an extension of a leading device platform but as belonging to that horizontal process of achieving the best outcome.

Do you have any use cases or customer examples you can share?

So we’ve been able to show that using these advanced platforms in the OR can lift performance level, and not only for surgeons but also for other physicians and clinical collaborators as well. For example, nurses. After the pandemic a lot of folks entered the nursing workforce without maybe as much training as they would have had before. And then there’s a lot of surgical volume right now because so many surgeries were bumped. So there are a lot of newer nurses who need to come up to speed very quickly. We’re increasingly deploying something like an interactive, step-by-step navigation guide in the OR. Getting step-by-step support in the right moment of the procedure can be extremely helpful to someone who may still be lacking confidence or experience in that setting.

How does Caresyntax work with partners to bring these platforms into ORs?

Being surgery specialists, we have a very good view of what the end applications and use cases should be, but we don’t have as much experience building the infrastructure. We don’t have the benchmarks and comparables from other use cases that may be similar in terms of the rigor and the actual architecture. And an integrated smart-surgery platform that is plug-and-play, that is very smart but not very heavy in terms of hardware content, something that is able to generate information but also has the capability and bandwidth to receive algorithm and produce AI and showcase it in real time—that’s a pretty sophisticated set of requirements.

Intel has been one of the partners that has really plugged in with us, almost inside our team, to make this happen. Designing the architecture, finding the right components, utilizing some of their components—such as OpenVINO^™ that allows for this AI penetration and usage—all of these things are very important. Without a partner like Intel we would have been, at the very least, much slower, looking for every piece ourselves and probably making more mistakes.

Alongside Intel, of course, we also work with cloud-solution providers—AWS and Google Cloud. Because there has to be an edge-to-cloud transition. As I mentioned before, it’s a preoperative, intra-operative, and postoperative space, so you have to continuously go to the edge and back to the cloud and make the information interchangeable. And actually they all collaborate in between themselves—Intel and Google, Intel and AWS—which has been very rewarding as well.

Of course, the pandemic was an impediment to innovation, but that has subsided lately. I think everybody’s really looking at surgery and saying, “It’s not as safe as flying; it’s not as safe as even some other medical procedures. It’s time to improve it.” And it takes an ecosystem of players to achieve that.

What’s your most important takeaway about the use of AI in surgery?

I very often see that folks think of surgery as something that’s been figured out, something that’s reached maturity and doesn’t require innovation. It doesn’t give me any pleasure to say that this is not the case. But there is the opportunity to get surgery to the same place as, say, aviation. I don’t think you and I would accept getting on a plane with a 15% chance of something going wrong in that flight.

It’s a huge problem that has not only clinical implications but cost implications. Next to pharmaceutical therapies, surgical therapies are the second-most-used way of correcting a disease. It’s probably 20%, 30% of all of healthcare spend in the US.

So if we’re going into a surgery, I think we should have the feeling that everything is going to be okay. And that should be backed by real statistics. We really can make surgery safer and smarter. It will have broad impact on patient health for millions of people, and a broad impact on cost as well. There’s ample room for improvement as long as the mindset for innovation is there.

Related Content

To learn more about AI-assisted surgical technology, listen to Staffing AI in the OR: With Caresyntax and follow Caresyntax at @caresyntax and on LinkedIn.

This article was edited by Erin Noble, copy editor.

From MRIs and CT scanners to electrocardiograms, oximeters, and blood pressure monitors, today’s medical machines collect an enormous amount of data about patient health. But getting that information into the hands of hospital doctors and nurses who need it is often a struggle.

One problem is that medical machines and devices are made by a wide range of manufacturers. As a result, the information they produce is scattered in incompatible formats across multiple data bases and IT systems, making it hard for doctors and nurses to assemble a coherent picture of patient health.

Another issue is hospitals’ restrictive hardware requirements. Nursing stations, operating rooms, and intensive care units—as well as the labs and pharmacies they connect with—must use computers that meet rigorous hygienic standards. And to successfully transmit vital patient data, the computers must also be very fast, reliable, easy to use, and secure.

Modern medical Panel PCs are designed to meet these challenges. They support data integration from a diverse range of medical equipment via an all-in-one, compact computer that meets hospital sanitary, usability, security, and reliability requirements. And they have the computing power to run the sophisticated medical software that provides caregivers the data they need not only to respond to emergencies but to consistently measure patient progress.

Unifying Patient Monitoring Solutions

Because healthcare technology continuously advances, most hospitals contain a wide range of medical equipment makes and models.

“Hospitals modernize their technology in phases, so they have a very heterogeneous hardware and software environment. Communication protocols are a big pain point,” says Guenter Deisenhofer, Product Manager at Kontron AG, a manufacturer of IoT equipment.

Machines that don’t intercommunicate erode procedural efficiency and make treatment decisions difficult—a problem Deisenhofer recently experienced firsthand after taking his son to a local hospital. First, an intake worker measured the boy’s heart rate, blood pressure, and oxygen level, recording the information on a slip of paper. His son was later seen by a doctor, who took his vital measurements again, writing them on a different paper before sending him off for X-rays—where the process was repeated yet again.

“At the end of the day, there were probably five pieces of paper. They had not continuously monitored his condition and nobody had an overview of it,” Deisenhofer says.

Fortunately, the situation turned out not to be serious. But with vital information arriving from different devices at different times—whether it’s recorded on paper or encoded in incompatible software programs—doctors can never be sure of what they might miss.

Medical edge computers, such as Kontron’s MediClient Panel PC, close the information gap, using a standard protocol to integrate data from machines, wearables, and patient health records. The Panel PC satisfies hospitals’ strict sanitary regulations and is readily accessible to caregivers, even if they’re wearing gloves. It conveys information from patient monitoring machines through wired or wireless connections to hospital communication hubs, such as nursing stations. High-performance Intel^® processors enable the monitoring machines’ software to run near-real time analytics on incoming results, helping doctors and nurses see patterns and spot anomalies that may suggest a diagnosis, or point to the need for specific tests.

“With continuous monitoring data available, doctors aren’t just reacting to an emergency. They can see, for example, if the heart rate is dropping and recovering over time. It helps them make better diagnostic and treatment decisions,” Deisenhofer says.

Monitoring can continue after patients are released from the hospital, with wearable devices seamlessly transmitting their information to the MediClient, where it can be integrated with patients’ previous records.

With hospitals increasingly relying on advanced #medical equipment, machine #manufacturers must be vigilant in keeping them updated, both to enable new #IoT functions and to guard against the latest cyberthreats. @Kontron via @insightdottech

Improving Machine Life Cycle Management with Medical Panel PCs

With hospitals increasingly relying on advanced medical equipment, machine manufacturers must be vigilant in keeping them updated, both to enable new IoT functions and to guard against the latest cyberthreats. Making these changes often involves not only software but firmware and sometimes hardware. That means devices like the MediClient PC must also be updated to continue providing hospitals the machines’ vital data.

As technology innovation accelerates, machines that were built to last 10 or 15 years often require several major updates. “Machine life cycle is getting more and more difficult to manage,” Deisenhofer says.

Kontron works closely with medical equipment manufacturers to keep up with planned changes and is often included in early product planning cycles. Because every hardware change requires extensive testing and recertification, close communication saves time. Manufacturers can get their products recertified in one go, instead of having to do so again after making modifications. Kontron also does third-party software installation and electrical testing of manufacturers’ equipment to help them resolve potential problems before release.

Collaboration with Kontron allows manufacturers to deliver upgrades to hospitals sooner—and hospitals can integrate the new capabilities into their medical systems as soon as they are available.

Bringing New Capabilities to the Edge

Working together, OEMs and Panel PC makers can extend the value of monitoring machines as technology advances. The more data the machines collect, the more system builders can improve their AI software, reducing false alarms and pinpointing problems. “For example, you could use AI to create warning scores for recognizing conditions well in advance of a critical situation,” Deisenhofer says.

And the processing speed of edge Panel PCs will improve, helping caregivers respond to identified health threats sooner.

Benefits like these suggest a bright future for patient monitoring machines and the medical Panel PCs that connect their data to doctors and nurses. As Deisenhofer says, “Our device cannot make decisions. But by bringing all the data together at the edge, it can help doctors make the right decisions.”

Edited by Georganne Benesch, Editorial Director for insight.tech.