Health and Life Sciences at the Edge: Recent Episodes
intel
The rising cost of health care has hit Americans differently depending on their economic status. There is no shortage of news stories covering the reality of those with higher incomes having higher life expectancies than those in lower financial brackets. Healthcare services such as clinical diagnostic testing may be out of reach for those who could really use it. So what are some the things a company can do to make clinical diagnostic testing more accessible for everyone? How can an important medical test be affordable to all?
On this episode of Health & Life Sciences at the Edge, host Michelle Dawn Mooney talks with Intel Head of Business Innovation and Life Science Specialist Blake Harlan and SiPhox Chief Product Officer Mike Dubrovsky about clinical diagnostic testing, the way it is used, and the way its use can be improved.
“Clinical diagnostic testing is actually a major part of the [healthcare] system,” says Dubrovsky. “Seventy percent of all medical decisions require diagnostic tests. However, the way that it’s done now is very reactive, so the whole system is based on first the patient having symptoms and then they’re put through a series of tests to try to resolve the problem.” Harlan and Dubrovsky further mention how it’s easy for those in higher income brackets to get tested more often, as those types of tests cost much more than the ones for reactive patients who are presently experiencing symptoms.
“The way the healthcare system works now is it’s really set up to, for the person, once they get sick, to deal with that” says Dubrovsky. He points out how the system is primarily designed to treat reactive care. “The problem is that there is a massive rise in healthcare costs, accompanied with actually a decrease in life expectancy, at least in America”, where six out of ten Americans have at least one chronic disease. In a country that spends so much on healthcare research, Harlan and Dubrovsky think the country needs to shift its thinking and policies when it comes to the health of its citizens.
Learn more about network operations center solutions by connecting with Blake Harlan and Mike Dubrovsky on LinkedIn or visiting SiPhox or Intel Health and Life Sciences.
Subscribe to this channel on Apple Podcasts, Spotify, and Google Podcasts to hear more from the Intel Network and Edge Solutions Group.
Healthcare organizations generate massive amounts of data, so much so that the challenge becomes how and where to move it and store it.
Health and Life Science at the Edge host, Gabrielle Bejarano, spoke with Zettar's Chin Fang and Intel's Michael McManus for a peek inside the technology solutions Zettar and Intel are partnering on to advance the challenge of data movement, processing, and storage for healthcare organizations.
Genomics files are a perfect example of the type of sizable data files that life science companies are creating. "When you go to genomics, for example, genomics files, a whole genome for a single person is about 350 gigabytes," McManus says. "So, if you're sequencing many people, you multiply however many people, you're sequencing times 350 gigabytes for your storage planning purpose." And he says cryoelectronic microscopy data sets are multiple terabytes in sizes. The data is broken into single gigabyte movies, and the average microscope can create thousands of those an hour. And then, the data requires analysis.
Chin adds there are some institutions Zettar works with whose processes generate files as big as four terabytes with rates of one terabyte per second that requires aggressive data reduction to reduce that to the gross rate of 800 Gbps, or 100 gigabytes per second.
These huge datasets cause a myriad of problems for technicians, scientists, and IT staff, including wasted time, frustration, and a loss of productivity.
Standard data-moving practices can't handle these complex situations, so Zettar's approach utilizes an efficient unified data-mover model. "With Intel's help, we developed this real-time capability, so now we make the real-time data movement readily available to anyone who needs it, and then this will have great benefits to accelerate the healthcare and life sciences research efforts going forward," Chin says.
Learn more about Health and Life Science data movement solutions by connecting with Chin Fang and Michael McManus on LinkedIn or visit Intel Health and Life Sciences.
Subscribe to this channel on Apple Podcasts, Spotify, and Google Podcasts to hear more from the Intel Health and Life Sciences at the Edge.
Medical imaging, an integral part of the cancer screening process, continues to experience rapid, year-over-year growth. Unfortunately, available radiologists cannot keep pace with the demand for their services fueled by the continued growth in cancer screening. New solutions are in desperate need to support diagnosis and improve workflows. Can AI play an important role in cancer screenings today and in the future? That’s what Health and Life Sciences at the Edge’s Tyler Kern wants to find out, and Intel’s Business Development and Sales Lead, Ryan Kim, and Lunit’s Director of INSIGHT Marketing, Jonathan Yang, have answers.
The disparity between the growing need for cancer screenings and the radiologists available to do the work can lead to missed early detection of cancers on chest X-rays and mammograms. Yang says this situation can create a big problem. “Regarding mammography, a high number of screenings can be read as false-positive, meaning that only a single digit percent of those recalled are retested,” he explains. “So, this is where AI can come in. AI isn’t here to replace radiologists, but it is here to support them, especially regarding cancer screening.” Yang sees many potential benefits AI can provide in assisting cancer screenings for patients, physicians, and medical institutions.
AI-driven solutions can reduce physicians’ reading time. Yang states, “This allows physicians to spend more time on the hard and tough cases. It also enables early detection of disease.”
Integrating AI into the cancer screening process is a seamless process today. Yang says there are two ways to integrate AI into the diagnostic imaging workflow. The first is when an image is acquired, and the second is during the image interpretation process. In both cases, Luni partners with companies with imaging platforms to make the integration seamless for healthcare institutions.
The utilization of Intel’s OpenVINO™ Toolkit assists in processing images quickly and minimizing the time from X-ray imaging to diagnosis. With OpenVINO™ technology, diagnostics can run on a CPU-only mode of operations, which reduces the cost of operations while delivering on the imaging speed physicians need.
Learn more about AI-enabled cancer screening solutions by connecting with Ryan Kim and Jonathan Yang on LinkedIn or visit Intel Health and Life Sciences and Lunit’s websites.
Subscribe to this channel on Apple Podcasts, Spotify, and Google Podcasts to hear more from the Intel Network and Edge Group.
Making sense of health data, telemedicine, and adopting AI/ML are some of the newest innovations in healthcare. Intel’s Nathan Peper, Head of Strategy and Business Innovations, and Patrick Boisseau, Director General, Strategic Initiatives at MedTech Europe, provided their insight to Michelle Dawn Mooney, host of Health and Life Sciences at the Edge, about the related policy challenges to powering the healthcare network in the EU & U.S.
Navigating interoperability in the U.S. is challenging. Currently, there is a large focus on lowering the associated costs of healthcare while improving the patient experience. With a shortage of healthcare workers, this proves difficult. Additionally, the hospital networks must remain business-oriented, pushing profit over expense.
“Policy is just the first step of admitting a problem we need to address, once you have these emerging innovations, it makes it a lot easier to push beyond the boundaries of policy to better benefit the healthcare system – patients & healthcare workers,” says Peper.
By expanding broadband to more rural areas of the US and increasing data storage and access regulations, federated learning can assist and potentially provide robust statistical model for medical data, while keeping patient privacy.
In Europe, some of the challenges are similar to the U.S. interoperability of systems. Yet, Europe comes with its’ unique challenges, including implementing a standardized solution for data transportation, support, and language translations across 27 embassies while maintaining internal communication.
“Policy is also an enabler, even though sometimes it seems like a constraint… digital technologies is a fantastic enabler and will sooner or later improve the delivery of healthcare, in central facilities, but also at the patient’s house. The biggest impact is the change of the relationship between the patient and healthcare professionals…Now because of digital technology, patients are empowered like never before,” explains Boisseau.
Currently, European regulations include the Artificial Intelligence Act, Data Act, and Cybersecurity, which help improve the adoption of digital technologies and privacy in Europe. Patients now have more control over information and their conditions while also having the ability to access a much larger knowledge set. The benefits of digital solutions include assisted diagnostic decision support systems at all levels and improved customer and patient-centric models.
“We are far from deployment, but the existing examples are very promising, and if we go towards a wider adoption, cost will go down and digital health is also one way to reduce cost-pressure on hospitals and one way to contribute to managing shortage of skills, “ says Boisseau.
Learn more about the policy challenges that are defining healthcare in the EU and U.S . by. Connecting with Nathan Peper and Patrick Boisseau on LinkedIn or visit:
MedTech Europe
Intel
Subscribe to this channel on Apple Podcasts, Spotify, and Google Podcasts to hear more from the Intel Internet of Things Group.
Intel’s Abhishek Khowala, principal health AI engineer, and Séverine Habert, AI engineering manager, discuss some of the enhancements in brain tumor segmentation for enabling diagnosis.
While most brain tumors are benign, early detection is critical for the best treatment options and outcomes. Assessing a diagnosis starts with MRI 2D and 3D imaging. Segmentation of the brain tumor – or separating the tumor from normal brain tissues – is essential to identifying three key factors to allow doctors to move forward:
- Is the tumor benign or malignant?
- The approximate tumor size and location.
- Plan out the treatment options.
“We need to segment out the tumor from the rest of the tissues around it,” Khowala says. “For that, there is the unit model. And that architecture works with fewer amounts of data yet provides a clearer segmentation result.”
The brain tumor segmentation (BraTS) combined with OpenVINO™ toolkit could optimize MRI results during tumor detection and monitoring. “Since this is something that has to happen worldwide, we need to deploy it at scale,” Khowala explains. Scaling requires overcoming a few challenges. Utilizing OpenVINO erases issues of high-cost GPU required for deploying AI solutions or perceived performance limitations of common frameworks such as PyTorch or TensorFlow. “Brain tumor segmentation is a perfect example of applying the most common architecture and using it for multiple devices from edge to handheld devices,” Habert adds.
For optimized AI, the provided data must be robust, which is not an easy task. According to Khowala, Expert radiologists are required to interpret the MRI images to get to the ground truth data. BraTS helps predict results and compare accuracy with provided ground truth results using the Sørensen–Dice coefficient datasets. Once the data is available, modeling can take place and assist medical professionals in their diagnosis.
Learn more about brain tumor segmentation solutions by connecting with Abhishek Khowala and Séverine Habert on LinkedIn or visit:
https://www.intel.com/content/www/us/en/healthcare-it/healthcare-overview.html.
Subscribe to this channel on Apple Podcasts, Spotify, and Google Podcasts to hear more from the Intel Internet of Things Group.
Healthcare faces many global challenges. There is a critical shortfall of skilled workers in diagnostic imaging that requires innovative solutions to help mitigate the gaps. An aging population requires care, which drives up diagnostic imaging needs but there are not enough workers to meet the demands. Intel is partnering with companies like MITIS Health to meet these challenges. Intel IoT Group’s Maria Meriacre, and MITIS Health’s CEO, Rahul Mehta, spoke with Gabrielle Bejarano on innovative ways to solve some of the issues through home diagnostic reporting.
Hiring more skilled workers seems the obvious answer, but if the past two years are any indication, finding and employing those skilled workers isn’t simple. After the toll of the pandemic, many healthcare workers are looking for different careers or shortening their hours. But if there is a silver lining to the pandemic, it could be adopting new technologies that can assist with these staffing challenges.
Remote work capabilities are one example. “The same applies to the healthcare environment,” Meriacre says. “We’ve seen a rise in what some would call telehealth.” Telehealth opens new avenues to seek care from professionals who no longer need to be in the exact location where a patient is, “From a telehealth point of view, looking at how technologies can enable more care where it isn’t possible to be face-to-face and getting a diagnosis in a timely manner,” notes Meriacre.
Taking these technologies, a step further, a professional could provide a radiology diagnosis through telehealth with an at-home diagnostic kit. “When you are looking at care equality, and you start thinking about how people are interacting and getting more access to being looked at from a patient perspective, the industry’s moved forward leaps and bounds.” Mehta says, “Trying to bring care to more remote places as well to help with care inequality has gotten more attention now that we have more telehealth solutions coming online.”
MITIS’s health virtual desktop infrastructure (VDI) provides at-home reporting specifically designed for diagnostic purposes. “We looked at the premise of VDI and the mechanics of how that was being delivered,” Mehta says, “We looked at the components available in the market today to see if there was a way of using commodity technology, hardware, and software to deliver that mechanism to a hospital and a radiologist.” There wasn’t a solution available, so MITIS built one through partnerships with Intel and others.
Learn more about health virtual desktop infrastructure by connecting with Maria Meriacre and Rahul Mehta on LinkedIn or visit:
Intel Health and Life Sciences
MITIS Health
Subscribe to this channel on Apple Podcasts, Spotify, and Google Podcasts to hear more from the Intel Internet of Things Group.
The importance of providing patients with a fast and accurate diagnosis cannot be overstated – it can be the difference between life and death. Unfortunately, the healthcare system is overwhelmed, and physicians cannot keep up with the overflow of patient needs in a timely manner.
Dr. Chai Xiangfei, CEO and co-founder at HY Medical, and Beenish Zia, chief architect of medical imaging at Intel, sit down with podcast host, Tyler Kern, to discuss the current challenges of providing a patient diagnosis in a timely manner and how artificial intelligence (AI) solutions can help.
Beenish discusses and highlights the complexities of diagnosing patient illnesses. “Medical diagnosis is a complex process which requires clinical skills and the need for clear decisions. And it needs to be balanced with acceptance for the ambiguity of many clinical situations,” Beenish explains.
According to Beenish, there are five major obstacles that delay diagnosis:
There are variations in individual logic and processes.
Because the nature of the evidence is not clear, further tests/screenings are required.
Standards vary globally, which causes longer diagnosis times.
Lack of evidence and data from diverse populations leads to not having good data sets to represent different communities.
The number of cases a physician must handle grows daily.
“For example, in certain hospitals in New York, we have heard that the radiologists get more than 1400 scans per day, and maybe there is just like a handful of radiologists to look at them. So, just the huge imbalance between the number of cases requiring diagnosis and the medical staff available to provide the diagnosis is not good,” Beenish explains.
To address this need, Dr. Chai talks about the role of AI in reducing diagnosis time to improve patient outcomes. “There are a huge amount of images created every day. In the meantime, actually, there’s a shortage of radiologists to read images. So, the real situation is that because there is a long waiting list of imaging… it can take hours or even days” for patients to get the result of their tests, notes Dr. Chai.
With traditional processes, scans are taken and added to a waiting list to be read and interpreted by radiologists, which can take up to 15 minutes. Current AI technology can give initial results within one minute and can ultimately reduce reading time by 30%. When radiology staff and doctors get the initial results, they can then quickly review the areas highlighted by the AI and then determine a path forward for diagnosis and treatments. Having this reduction in reading time, is helping patients get treated sooner.
To learn more about how AI can transform the diagnostic process and improve patient outcomes, connect with Dr. Chai Xiangfei and Beenish Zia on LinkedIn.
To find out more about HY Medical please visit: https://www.huiyihuiying.com/#page1.
Learn about various technologies Intel offers to advance the healthcare field — from compute to storage and networking to AI — visit: https://www.intel.com/content/www/us/en/healthcare-it/healthcare-overview.html
Subscribe to this channel on Apple Podcasts, Spotify, and Google Podcasts to hear more from the Intel Internet of Things Group.
Ultrasound analysis leads the charge in advancements in diagnosing and treating cardiovascular disease. Intel’s Yehudit Levi and DiA Imaging Analysis’ Hila Goldman-Aslan joined Michelle Dawn Mooney for insights on the AI and technology advancements in ultrasound that provide better diagnosis and treatment of cardiovascular disease.
Hila Goldman-Aslan says that early detection is the go-to strategy to prevent a significant cardiac incident or death. “And ultrasound is the leading modality to identify those abnormalities in the heart, and it’s been increasingly used in all kinds of settings. You can see an ultrasound and cardiac ultrasound being used at the point of care in the emergency rooms and the ICU.”
A 2020 article from the American College of cardiology notes a rise in cardiovascular disease. With one-third of global deaths in 2019 attributed to cardiovascular disease, it is critical to find preventative solutions to reduce these alarming numbers. And this situation is where AI can play an essential role. “Artificial intelligence is all about data,” says Yehudit Levi. “The health segment is unique in the sense that a lot of medical records have been collected through the years. Artificial intelligence software enables processing these masses of data.” The insights gleaned from all this data and knowledge, once made accessible to the medical staff, help highlight abnormalities that might otherwise prove difficult to detect.
DiA Imaging Analysis powered through Intel technology strives to allow for more efficient cardiovascular results and better treatment. “We see physicians looking for more automated procedures and processes to help them make better decisions based on more objective information,” Goldman-Aslan says. Until recently, the review of ultrasound images relied solely on the operator, which creates a layer of subjectivity. With AI analysis, the process becomes more objective and provides better information for the cardiovascular team to make decisions. The increase in cardiac procedures and a shortage of staff analysts available make DiA Imaging ultrasound analysis solutions integral.
“Improving the diagnostic time is super important,” Levi stresses. “It allows more patients to be diagnosed.” Together, DiA and Intel make these ultrasound capabilities scalable for healthcare.
Learn more about Health and Life Sciences by connecting with Yehudit Levi and Hila Goldman-Aslan on LinkedIn or visit:
Intel Health and Life Sciences
DiA Imaging Analysis
Subscribe to this channel on Apple Podcasts, Spotify, and Google Podcasts to hear more from the Intel Internet of Things Group.
Staffing shortages in healthcare is a major corner, one department particularly affected is baby labor and delivery, where new parents rely on optimal care. While the birthing process is a natural occurrence, in today’s healthcare world, providing the best medical advancement possible in these situations is the goal, and technology can help physicians get there.
Intel’s Chief Healthcare IoT Solution Architect, Karen Perry, joins her identical twin sister Kelli Parker, a nurse practitioner, to discuss these issues with host of Health and Life Sciences at the Edge, Michelle Dawn Mooney.
Intel and its partners created solutions to observe the patient remotely through two-way video and audio cameras. This approach allows a nurse to triage a situation even if they cannot be where the patient is when a problem occurs.
With trying to ensure the first few moments and days of a new child’s life goes as smoothly as possible, there are clearly many reasons and needs for advanced monitoring during the labor process of birthing. Some, Parker says, are for medical concerns for the patient and baby, but others focus on how the mother chooses to bare that child. No matter the method —natural versus cesarean and with or without an epidural — the better the monitoring system, the better the outcome.
Another exciting advancement in labor and delivery practices is the introduction of AI to alert a medical professional if an issue arises. “We’re working on algorithms to detect the patient’s motion,” Perry adds. “Also, to know what’s a normal motion or an abnormal motion.”
To make these technology advancements possible, Intel works with clinicians to determine their needs and where technology can play a role that makes sense in the overall healthcare workflow.
This technology works in two ways:
Monitoring patients “behind the scenes” for when things are going right, giving physicians and family peace of mind, and
Alerting clinicians when there is a concern and immediate action is needed.
In the case of mothers and babies during labor, delivery, and recovery, a seamless workflow is essential.
Learn more about Intel’s solutions by connecting with Karen Perry and Kelli Parker on LinkedIn or visit:
Smart Hospital Business Brief.
Intel Internet of Things Group
Subscribe to this channel on Apple Podcasts, Spotify, and Google Podcasts to hear more from the Intel Internet of Things Group.
The healthcare industry is undergoing dramatic changes thanks to rapid digitization and the adoption of AI. Nowhere are these changes more evident than in hospitals, where increased connectivity, data analysis at the edge, and patient monitoring are helping staff rethink patient care. In this episode of “Health and Life Sciences at the Edge,” host Daniel Litman talks with Karen Perry, Intel’s Chief Healthcare IoT Solution Architect, and Lisbeth Votruba, Chief Clinical Innovation Officer at AvaSure, about how smart technologies can improve care, particularly when implemented in patient rooms.
Karen Perry, whose job is to look at all parts of the Smart Hospital puzzle, says “We have to work closely with clinicians to make sure that when we introduce a technology it’s something they can leverage.” Lisbeth Votruba agrees saying, “I believe we’ve reached an inflection point we’ve been heading towards for years. More and more nurses are considering leaving and more than 28% of new nurses will leave the profession within the first year.”
The good news, according to Votruba, is that there’s a new urgency and willingness to invest in smart solutions that support the workforce. “One of the big barriers to implementing forward thinking technology in hospitals is competing priorities,” says Votruba. “I’m seeing more and more leaders prioritizing technologies that help educate, retain, and make the work of new staff easier.”
Perry and Votruba are excited about the benefits of adding smart technologies to patient rooms, particularly computer vision and audio. However, Perry is quick to point out that the three metrics of success – positive outcomes for patients, the workforce, and the hospital – must be met. “As we introduce capabilities into hospitals, we have to create measurable goals,” she says. “Those three metrics are like the three legs of a tripod. We have to hit every one of them for hospitals to invest.”
Looking to the future, Perry and Votruba see the use of smart technologies as a journey. “What can happen next,” says Perry, “is we follow patients on their journeys through the system. That will allow us to create a more connected, cohesive, and effective healthcare system.”
I’m very hopeful,” says Votruba. “I’m not afraid of technology. I’ve seen the benefits and believe that the investment we’re making in technology will improve both the care patients receive and the humanity of how it’s delivered. Hopefully smart tech will let caregivers get back to the joy of why they chose this profession.”
Connect with Karen Perry and Lisbeth Votruba on LinkedIn or visit:
Intel Smart Hospitals
Intel Health and Life Sciences
AvaSure
Subscribe to this channel on Apple Podcasts, Spotify, or Google Podcasts to hear more from the Intel Internet of Things Group.
Telehealth has been around for quite a while. However, it took the pandemic to generate widespread adoption and acceptance of telehealth in hospitals globally. In this episode of “Health and Life Sciences at the Edge,” host Justin Honore talks with Intel’s Ed Buffone about the technologies that are making Smart Hospitals a reality and the challenges that must still be addressed.
“During the pandemic, the healthcare system was under siege. This virus was draining the resources of every hospital in the country,” says Buffone. “We needed alternative ways to reach and treat patients. The two technologies that made that possible were virtual visits and remote monitoring.” However, telehealth encompasses a broad range of technologies and applications that must work seamlessly together to make Smart Hospitals viable. “What hospitals lack is an organized, holistic approach to telehealth,” says Buffone. “In many hospitals, individual departments deploy telehealth, but they’re siloed. The solution is a cohesive, enterprise-wide approach.”
The three major technologies driving the move to Smart Hospitals are Artificial Intelligence (AI), Analytics, and 5G. “AI is a key element that makes hospitals Smart Hospitals,” says Buffone. “Healthcare is an industry that is rich in data, and AI takes advantage of that data to give clinicians more insight into patient care.” Buffone refers to AI as Assisted rather than Artificial Intelligence. “When it comes to healthcare, there’s nothing artificial about AI,” he says. “The clinician is making the intelligent decisions for the patient. AI just provides data that can help.”
Working alongside AI are analytics and 5G. Analytics provide insights into patient populations and 5G extends patient care networks, making it possible to follow patients through their care journeys.
According to Buffone, there are three main hurdles to implementing these technologies in hospitals – finances, staffing, and lack of organization. “I think the real solution is maturity,” says Buffone. “Maturity of the technologies and the processes. As these technologies become more accepted and integrated into the workflow, the overall financial lift lessens. At that point everyone has a clear understanding of the return on investment and how the technologies provide for lower costs, most efficiency, and better patient satisfaction. Ultimately, that’s what we’re after,” Buffone adds. “It’s all about the patient journey, patient satisfaction, and patient safety. If you can do that while lowering costs and providing more operational efficiencies those barriers will fall away.”
Connect with Ed Buffone on LinkedIn.
Follow us on Twitter: @IntelHealth
To learn more about telehealth and Smart Hospitals visit: https://www.intel.com/content/www/us/en/healthcare-it/smart-hospital.html
Subscribe to this channel on Apple Podcasts, Spotify, and Google Podcasts to hear more from the Intel Internet of Things Group.
Over the past two years, the healthcare industry has experienced significant changes in how, when, and where it utilizes technology. In addition, the way patients perceive and want to interact with the healthcare system has undergone a dramatic shift. In this “Health and Life Sciences at the Edge” podcast, host Michelle Mooney talks with Alex Flores, Intel’s Global Health Solutions vertical, about the causes and trends driving these changes.
“I believe three things are behind the changes we’re seeing,” says Flores. “The healthcare industry is moving at pandemic-speed; there’s been an accelerated investment in and adoption of artificial intelligence (AI); and patient expectations have evolved.”
The pandemic forced healthcare leaders to quickly digitize many aspects of the system so care could be delivered outside the traditional four walls of clinics and hospitals. This increased the adoption AI, with more than 84% of healthcare leaders surveyed saying they expect widespread AI adoption within five years.
Patients have come to expect healthcare to be more personal, convenient, and flexible. “The resulting ‘hybrid care models’ have created a demand for additional compute, security, data storage, and analysis, particularly in hospitals” says Flores.
Tech and data-driven hospitals (Smart Hospitals) are employing extensible tech frameworks that enable workflow agility across healthcare partners, providers, and systems. “In today’s digital-first world,” says Flores, “Smart Hospitals are significantly increasing connectivity, intelligence and automation. Ultimately, these lead to improved care models and workflow efficiencies.”
Intel’s superpowers – AI, ubiquitous computing, pervasive connectivity, edge to cloud infrastructure – are particularly relevant to healthcare due to the sheer amount of data generated by the healthcare system. “It’s estimated that healthcare produced about 30% of the world’s data produced worldwide,” says Flores, “but only 5% of that data gets analyzed.” This points to the need to enable more analytics at the “intelligent edge” in hospitals and clinics.
Looking to the future, Flores is excited about how digital technologies can revolutionize the healthcare industry. In particular, he points to the concept of the “digital twin,” which allows a comprehensive and predictive model of a patient. to be generated by aggregating and coupling data with real-time information. “The digital-twin can also be applied to physical hospitals and allows stakeholders to review how implementing innovative solutions would impact operational efficiencies.”
“The bottom line,” says Flores, “is that in order for the healthcare industry to truly transform, it will need to embrace technology and dedicate itself to improving the patient journey.”
Connect with Alex Flores on LinkedIn
To learn more about Smart Hospitals, visit Smart Hospital
Follow @IntelHealth on Twitter
Subscribe to this channel on Apple Podcasts, Spotify, and Google Podcasts to hear more from the Intel Internet of Things Group.
Heterogenous computer systems - systems that contain different kinds of computational units (CPUs, GPUs, etc.) controlled by a general-purpose processor (GPP) and augmented by accelerators (XPUs) - are here to stay. However, the computational complexity inherent in these systems creates some unique challenges in the healthcare industry. In today’s “Health and Life Sciences at the Edge” podcast, Intel’s Beenish Zia, Chief Architect for Medical Imaging in Intel’s Internet of Things Group, and Evgeny Drapkin, Chief Engineer for GE Healthcare Digital Platforms, talk with Tyler Kern about those challenges and solutions.
According to Drapkin, both the need for heterogeneous computing and the biggest challenges to its successful deployment can be illustrated by medical imaging. “In medical imaging, being able to deliver results from scans as quickly as possible is a necessity,” says Drapkin. “In many cases, like in stroke management, the speed of delivery can directly impact patient outcomes.” With the level of computational complexity growing every year, finding ways to increase image processing speeds is both imperative and challenging.
Enter heterogeneous computing using the Intel® oneAPI Toolkit. OneAPI is designed to meet the three biggest challenges developers and programmers face when creating heterogeneous systems:
- Determining which hardware architecture to use
- Selecting the proper software model
- Porting legacy software in ways that take advantage of modern technologies
“OneAPI stands for One Application Programming Interface,” says Zia. “It simplifies software development and programming by providing a unified programming model. This model gives programmers the freedom to select the best hardware for their workloads, optimizes hardware performance, and removes hardware vendor lock-in.” Best of all, the same learning model applies to all industries and helps build community and industry collaboration.
Both Zia and Drapkin agree that heterogeneous computing is needed to identify and map algorithms to accelerator devices, then program those devices to deliver faster results. “Industries like medical imaging are driving the need for a common programming language that can transform how software coding is approached,” say Zia. “OneAPI provides a unified programming model that can be applied to all industries. Heterogeneous computing is inevitable. It’s up to us to optimize it.”
Connect with Beenish Zia and Evgeny Drapkin on LinkedIn.
More about how oneAPI could assist in heterogenous computing and rapid software deployment can be found at oneapi.io.
Subscribe to this channel on Apple Podcasts, Spotify, and Google Podcasts to hear more from the Intel Internet of Things Group.
Over the last decade, AI-driven computer vision (CV) and machine learning (ML) have become ubiquitous. However, the major trends in both industries began more than two decades ago and continue to shape advances in technology and regulatory changes today. In this episode of Health and Life Sciences at the Edge, Stephanie Cope and Kaleb Kurther - Lab and Life Sciences experts at Intel – discuss the evolution of AI-driven CV and ML, the potential applications of CV and ML in hospital and clinical settings, and how Intel is creating both the hardware and software needed to realize their potential.
Within the last 2 years we have seen even more changes, specifically regarding robotics and automation options that are available to instrument makers and designers today. “Most recently, with the pandemic the ability to test accurately and at scale are paramount today,” Cope said. This includes concepts like the way basic chemistry tasks are performed. Automation like liquid handling robotics is now being integrated into these instruments today, vastly changing and improving the amount of throughput that these instruments can provide in a hospital and clinical setting.
Kurther, who has dealt with automated equipment in high-volume environments, sees a practical use for ML and CV. “Using machine learning can improve diagnostics and prevent errors,” he says. “A lot of times you’re dealing with an error after it happens. We call it coming upon the ‘scene of the crime.’ You ask what happened? How do I prevent this? How do I predict this? Machine learning and computer vision let you walk the timeline back to see exactly what happened.”
Looking five years into the future, both Cope and Kurther are excited about the possibilities. “On the application side, “says Cope, “we’re seeing how integrated genomics can create opportunities for personalized medicine and change how patients are treated.” For his part, Kurther says, “I am a big fan of Moore’s Law. I look forward to all we will be able to do with transistors of the future.”
To continue the conversation with Stephanie Cope and Kaleb Kurther, connect with them on LinkedIn.
For more information, please see Intel’s Business Brief, “Powering the Future of Automated Clinical Chemistry and Blood Bank System” https://intel.ly/3n7a7ne
To learn more about Computer Vision here: https://www.intel.com/content/www/us/en/internet-of-things/computer-vision/vision-products.html
Subscribe to this channel on Apple Podcasts, Spotify, and Google Podcasts to hear more from the Intel Internet of Things Group.
The successful use of AI for insurance claim processing depends on identifying and optimizing the performance of hardware and software tools that increase its efficiency, flexibility, and speed. Vasant Kearney, Ph.D., CTO of Retrace Labs, and Ravi Panchumarthy, Ph.D., Machine Learning Engineer at Intel Corporation spoke about the specific challenges faced by the dental industry when it uses cloud-based AI computing to process claims, and the creative innovations that Intel and Retrace have developed to address those challenges.
“The thing that makes AI-driven insurance claim processing challenging is the throughput,” Kearney says. “If you’re in a single hospital or small clinic, scalability isn’t much of an issue. But once you move into the insurance world - where the volume is much higher - you have to manage spikes in throughput.” The ability of the AI algorithms being used to manage these spikes (scaling performance up or down based on demand) is determined by how much compute has been allocated to solving the problem. This is where the choice of hardware and software becomes critical.
Most data scientists are familiar with GPUs and choose them when deploying models in production. But GPUs can be costly and create delays due to the way they handle memory and how they are deployed onto scalable tools. “It’s not trivial to share memory between GPUs,” says Kearney. “So, you’re limited by the rather low-memory footprint of each GPU. In AWS, you have GPUs in the range of 12 gigabytes. But CPUs can get up into the terabytes.” This means that many more models can be stored on each instance, making CPUs ideal for healthcare where many different models are often needed to make diagnoses.
Both Panchumarthy and Kearney are excited about the future of AI-driven cloud computing for the insurance industry. “There is great synergy between cloud computing and cutting-edge hardware and software solutions from Retrace and Intel,” says Panchumarthy. “All of these are helping drive even more intelligent and robust medical AI solutions. It’s an exciting place to be.”
Learn more about AI deployment solutions by connecting with Vasant Kearney and Ravi Panchumarthy on LinkedIn or visit Intel’s AI and Deep Learning Solutions to learn more about AI-driven Solutions: https://www.intel.com/content/www/us/en/artificial-intelligence/overview.html
To get started with OpenVINO: https://docs.openvinotoolkit.org/latest/index.html
Learn more about Retrace: https://retrace.ai/media-and-news/
Subscribe to this channel on Apple Podcasts, Spotify, or Google Podcasts to hear more from the Intel Internet of Things Group.
Artificial intelligence (AI) is often associated with consumer-facing and mainstream industries such as gaming, securities and retail, but it plays just as crucial of a role in the biopharma industry. Sandi Colner, head of lab and life sciences vertical for Intel, and Stephanie Davies, head of science for cell therapy at ValitaCell, joined Host Tyler Kern to discuss this.
“The reason these therapies are important is that they really are a potential solution for a lot of diseases that don’t have alternative treatments like certain types of cancer, or OA and Crohn’s disease,” Davies explained.
Currently, the manufacturing process is still crude. This is partly because tools to monitor cell health and function can take half the time the entire process takes. Embedding a new breed of simple, automatable analytics in these processes — such as image and AI based assessment of cell quality — can help produce a high yield of quality cells to treat patients. This can help decrease the strenuous manufacturing processes that limit the industry.
Sandi Colner, head of Lab and Life Sciences at Intel says that the biopharmaceutical industry is undergoing an immense digital transformation driven by AI. “Although the application of AI in manufacturing is still in its nascent stage,” Colner says, “it’s already created an extensive ecosystem of different AI solution providers.”
Davies believes the key to accelerating this lies in industry collaboration. “I think these amazing collaborations that we are having across various parts of the industry are going to help connect different siloed data sets. It’s going to enable us to have access to sufficiently large data sets that are diverse and that are really going to be needed to train these AI models,” she said.
Looking to the future both Davies and Colner see many ways AI can help solve problems in the emerging cell and gene therapy industry. “This is difficult science, and we need to acknowledge that,” says Colner. She continues to explain that the next collaborations happening across various parts of the AI industry will enable new cell therapies to be delivered.
If you are interested in learning more or collaborating in the transformation of the biopharma industry, please connect with Stephanie Davies or Sandi Colner on LinkedIn.
Subscribe to this channel on Apple Podcasts, Spotify or Google Podcasts to hear more from the Intel Internet of Things Group.
More and more of today’s medical imaging devices, such as CT, ultrasound, and MRI scanners, rely on real-time AI inferencing at the edge to make critical medical decisions while patients are being treated. Intel’s Deepthi Karkada, a deep-learning software engineer, and Ryan Loney, Product Manager for OpenVINO™ spoke to Hilary Kennedy about recent trends in AI-based medical imaging and how Intel and its partners are helping identify and address the rapidly changing needs of this burgeoning industry.
“Real-time medical imaging at the edge is important because it enables healthcare providers to get results from scans, run inferences, and make decisions about medical care at the patient’s bedside,” says Looney. “Often these results need to be obtained and processed in two seconds or less.”
Computing at the edge is not without its issues, however. Three of the major hurdles Intel and its partners routinely face are: limited memory in low-power devices, binary size, and latency. “Every megabyte counts when you’re deploying on low-power medical devices with limited memory,” says Looney. “Analytics need to be run in as close to real-time as possible.
“We know that AI and similar techniques are being adopted in the fields of medical imaging,” Karkada said. “These techniques include things like object detection and semantics segmentation. These techniques help radiologists quickly identify issues and result in many benefits. Many of our partners have been leveraging these advancements in these technologies.”
“Intel offers a portfolio of hardware solutions targeted for AI inferencing,” Karkada said. “This includes solutions like the Intel Xeon® processors, core processors, and FPGAs, that our partners have been able to leverage. On the software side, our OpenVINO™ Toolkit provides accelerated inferencing solutions. These also take advantage of the hardware features, so they’re tightly coupled and integrated.
Learn more about AI and edge solutions for medical imaging, and other health and life sciences, by connecting with Deepthi Karkada and Ryan Loney on LinkedIn, or read more about Intel’s medical imaging solutions online.
Learn how to optimize a CT model using OpenVINO here: https://github.com/openvinotoolkit/openvino_notebooks/tree/main/notebooks/110-ct-segmentation-quantize
Hear some of our customer success stories here.
Subscribe to the “Health and Life Sciences at the Edge” channel on Apple Podcasts, Spotify or Google Podcasts to hear more from the Intel Internet of Things Group.
The demand created by the pandemic for healthcare that could be delivered outside traditional office settings has resulted in the widespread adoption of Telehealth and challenged software and infrastructure providers to address key issues related to its delivery. “We’ve been promoting Telehealth for many years,” says Wendy Bohner, “but it took the pandemic for it to become mainstream – to become the new normal.”
According to Andrew Lamkin, Telehealth is not only efficient and effective, but popular as well. “Patients and doctors are liking it,” Andrew Lampkin says. “It is easier to get the care they need and clinicians are providing satisfactory levels of follow-up care. It’s really a good fit.”
-
Some of the additional benefits provided by Telehealth include:
-
Lower cost alternative: According to the American Medical Association and Wellness Council of America, “…almost 75% of all doctor, urgent care, and ER visits are either unnecessary or could be handled safely and effectively over the phone or video.”
-
Better utilization of clinical staff: Telehealth can free clinicians from day-to-day tasks and enable them to focus on skills they are uniquely equipped to provide.
-
Improved service experience: ability to diagnose and treat a patient virtually as compared to having the patient onsite in a waiting room for periods of time.
Some of the challenges of delivering remote care include:
-
Ensuring that adding Telehealth to clinicians’ workflows doesn’t make it harder for them to accomplish their jobs. “We can’t have silos,” Bohner says. “The elements need to be connected and carefully integrated so clinicians can be focused on their patients.”
-
Helping clinicians recognize visual cues while interacting with patients via video. Data analytics is starting to address this issue by spotting abnormalities and inconsistencies in patient behavior.
-
Ensuring that Telehealth solutions are reliable, secure, and scalable.
Intel is collaborating with the health ecosystem to address the challenges inherent in the delivery of Telehealth “So much of this starts with software,” says Lamkin. “We wanted to make sure that Intel’s array of software was available in pre-packaged, ready-to-use form that was right for getting started quickly with Telehealth.”
Both Bohner and Lamkin agree that the benefits of Telehealth far outweigh the challenges and encourage healthcare providers to look into making it part of their services. “Healthcare providers need to take advantage of these technologies because we can help patients stay healthy,” Bohner said. “Patient care can be delivered efficiently and effectively using remote technologies.”
To learn more, visit the Intel telehealth site at https://www.intel.com/content/www/us/en/healthcare-it/telemedicine.html and access the Intel Telehealth Remote Monitoring Reference Implementation here for prototyping. Wendy Bohner and Andrew Lamkin welcome feedback and communication via LinkedIn.
Subscribe to this channel on Apple Podcasts, Spotify, or Google Podcasts to hear more from the Intel Internet of Things Group.
In this episode of Health and Life Sciences at the Edge, host Tyler Kern speaks with Beenish Zia, an electrical engineer working as a platform architect in Health and Life Sciences at Intel, and Joy Yun, who interned in Health and Life Sciences at Intel, about the technological advances fueling the rapid expansion of the medical imaging industry.
The U.S. Food and Drug Administration defines medical imaging as several different technologies used to view the human body to diagnose, monitor, or treat medical conditions. Each type of technology gives different types of information about the area of the body being studied or treated. Examples include Ultrasound, X-ray, magnetic resonance imaging (MRI), and computer tomography (CT).
According to a market insights report by Research and Markets, the global medical imaging market size is expected to reach $28.6 billion by 2028. Zia attributes this rapid growth to advances in medical imaging hardware and software. “These advances include a variety of technical improvements,” says Zia, “from the devices that generate the raw data to how data is processed, stored, and transferred before it can be viewed by a radiologist or a healthcare technician.”
-
Three specific developments are driving advances in medical image processing: Convolutions and Cross-correlations: Software processes used to identify issues and refine, adjust, or modify image quality.
-
Parallel Programming: Allows the effective distribution of workloads over the computational resources available.
-
oneAPI Implementation: oneAPI is a cross industry, open standards-based, unified programming model that allows developers to write code in a common language, thereby enabling companies to code without having to learn three or more language constructs. “I believe oneAPI will provide developers with faster application performance, more productivity, and hopefully greater innovation effect,” says Yun.
According to Zia, another major advantage to using oneAPI, including Intel’s oneAPI implementation, is that it has the potential to end hardware-vendor lock-in. “Historically, when developers needed to move their application to a new hardware or target device based on a different architecture than what they were using, they would have to create an entirely new code base,” says Zia. “Those extra costs and delays are never welcome.” The goal of Intel’s oneAPI implementation is threefold:
-
Increase application portability
-
Raise developer productivity
-
Deliver peak performance to high-growth applications in data centers, at the Edge, and in the Cloud.
To learn more:
-
Connect with Beenish Zia on LinkedIn.
-
Take a look at these two white papers:
-
Fast Fourier Transform (FFT) and Convolution in Medical Image Reconstruction: https://www.intel.com/content/www/us/en/developer/articles/technical/fast-fourier-transform-and-convolution-in-medical-image-reconstruction.html
-
oneAPI for Healthcare: C++ to DPC++ Migration Example: https://www.intel.com/content/www/us/en/developer/articles/technical/oneapi-cpp-to-dpcpp-conversion.html?wapkw=beenish%20zia
-
Listen to our oneAPI in DevFest presentation with GE Healthcare: https://www.intel.com/content/www/us/en/events/developer/devfest-2021.html?videoId=6279852068001
-
If you are a developer interested in using oneAPI, visit “oneAPI: A New Era of Heterogeneous Computing: https://www.intel.com/content/www/us/en/developer/tools/oneapi/overview.html#gs.fgsjop
Subscribe to this channel on Apple Podcasts, Spotify or Google Podcasts to hear more from the Intel Internet of Things Group.
Medical imaging is pioneering the use of AI and analytics to improve the way healthcare is delivered. As the Head of Health Solutions for Intel’s Internet of Things Group, in the Health and Life Sciences and Emerging Technology organization, Alex Flores is uniquely qualified to talk about how AI and Edge technologies are improving the accuracy of medical imaging analysis. This trend began about ten years ago and has accelerated rapidly over the past three years.
“In July of 2020, Intel conducted a survey of over 200 US healthcare leaders. We were looking at key technology trends and the way technology needs have changed post-COVID-19,” Flores said. “We learned that about 45% of the respondents were using, or planning to use, AI in 2020 before COVID-19. That percentage jumped to about 84% since the onset of COVID-19.”
Flores discussed the benefits and problems associated with integrating Edge technology and AI into medical imaging. While adding more compute to or next to devices can shorten the time needed for clinicians to do their jobs, it can also create challenges for manufacturers of Edge devices.
“Edge computing is about doing more processing on or next to a device,” said Flores. “But when you add an external device, power consumption and acoustics become an issue. You can’t just stuff more power-hungry GPUs into an Edge device because they will increase fan size and noise. Imagine sitting in a clinician’s office reviewing your results but not being able to hear because the Edge server under their desk is too loud.”
It’s estimated that the health and life sciences industry produces about a third of the world’s data. But less than three percent of that data has been analyzed. Flores believes that by partnering with health and life sciences industry leaders to address the challenges of harnessing these vast amounts of data Intel can help them come up with actionable insights.
“In my opinion, we haven’t even scratched the surface of what is possible,” Flores said. “We work with our partners to architect optimum hardware platforms, then layer the right software tools to ensure that the solution is optimized to meet their needs. The exciting promise of AI in medical imaging is that it has progressed significantly and in my opinion is continuing to accelerate.”
Learn more about improving the accuracy of medical imaging analysis with AI by connecting with Alex Flores on LinkedIn or visit http://www.intel.com/healthcare
Subscribe to this channel on Apple Podcasts, Spotify, or Google Podcasts to hear more from the Intel Internet of Things Group.
The American Journal of Medicine noted that healthcare organizations, or HCOs, require a framework that preserves privacy and supports the kind of data collaboration that will make tomorrow’s scientific and clinical advances possible while still supporting improved patient outcomes and experiences. Establishing this framework is challenging. Can AI solutions meet those privacy requirements for healthcare and life science applications? This is the question host Hilary Kennedy posed to BeeKeeperAI’s CEO Michael Blum, MD and Intel’s Health and Life Sciences General Manager, Chris Gough.
“Even large health organizations do not have the diversity, quality, or amount of data required to create high-quality, generalizable AI,” said Blum. “The solution is a federated framework that allows each data steward to ensure the privacy of their data by keeping it in their protected cloud, while validating and training AI across all of those stewards.”
Gough agreed, saying, “If you can reduce the friction to accessing large, diverse, high-quality data sets - and use those data sets as part of the AI training step - then the algorithms that result will be more accurate for multiple populations.”
“Healthcare AI is projected to be a $46 billion dollar market. When you turn that into human impact, there’s a tremendous opportunity to improve the quality of care and reduce the cost of care,” said Blum. “But the challenge we see over-and-over again is that the algorithms don’t generalize across populations. Their accuracy and performance fall off dramatically. We have to get rapid access to much broader and much more diverse data sets.”
“We really have a supply and demand problem in healthcare,” Gough added. “AI will help the industry better target the scarce resources it has, allowing organizations to be less reactive and more proactive and predictive.”
To learn more about Michael Blum, MD, and Chris Gough:
- Connect with them on LinkedIn
- Visit https://www.beekeeperai.com/
- Read the “Privacy-Preserving Data-Collaboration Methods that Accelerate Healthcare Innovation” white paper: https://www.intel.com/content/www/us/en/healthcare-it/resources/confidential-computing-whitepaper.html
Subscribe to this channel on Apple Podcasts, Spotify, or Google Podcasts to hear more from the Intel Internet of Things Group.