Predictive analytics can save lives. Healthcare has always come with the fine balance of prioritizing care and allocating scarce resources. Providers are faced with tough decisions about how long doctors or nurses see patients, who occupies their limited number of hospital beds, or how to distribute precious resources like donated organs.
If providers could simply divide healthcare evenly among the population, matters would be simple. But different people need different treatments, levels of care, and services within healthcare. Allocating resources in a hospital or clinic in fair, moral, and equitable ways remains one of the industry’s most complex puzzles.
Healthcare, Data, and AI
Eighty-six percent of the nation’s $2.7 trillion annual healthcare expenditures go toward people with chronic and mental health conditions. The sickest 5 percent of patients consume 50 percent of the costs.
Healthcare providers look for ways to prevent these patients from falling into acute, high-cost situations. One way to do that is to make care proactive instead of reactive, and intervening at the right time before a condition worsens to a point where it affects costs, care, and patient well-being.
Healthcare is known for having large, varied, and comprehensive data sets. Patients fill out forms at the start of each visit, nurses chart during each shift, and every test scheduled by a physician generates an abundance of insights. This data is valuable on a micro level for patients and their care providers, and on a macro level for researchers who study public health, epidemiology, and healthcare policy.
AI offers the means to leverage healthcare’s data into targeted, preventive care for a community, reducing both the burden of chronic disease and the cost of care. Leading healthcare organizations are already using AI to sift through data sets and find previously hidden insights. Even in the past few years, the new technology has had a transformational impact on healthcare, and Intel technology is now a crucial part of how care providers are finding actionable insights.
AI Is Already Making a Difference
It’s not hard to find examples of how AI is improving healthcare outcomes today. At NewYork-Presbyterian Hospital, a new AI command center helps nurses in a remote control room monitor patients and processes with artificial intelligence tools to reduce alert fatigue and free up on-premises clinicians to spend more time with patients.
This has helped NewYork-Presbyterian decrease redundancy in tasks performed by registered nurses and doctors, reduce the number of team members physically required to monitor patients, and sizably cut down the amount of time staff spend inputting patient data.
In San Diego, Intel worked with Sharp HealthCare to use AI in identifying patients who were at risk of sudden decline, potentially requiring a rapid response team. The analytics work has shown predictability within 80 percent accuracy whether a rapid response call is likely to occur in the next hour. This enables Sharp HealthCare to optimize and proactively place rapid response teams at key points in hospitals, and even intervene before the situation becomes more life-threatening.
Intel and Our Partners
Our portfolio of advanced technology and AI is empowering healthcare innovators to create new, data-driven solutions that are dramatically reshaping the boundaries and precision of healthcare delivery.
We’ll also share open source predictive models and frameworks that are optimized to run on Intel Xeon Scalable processors in the data center to underpin this process, and showcase Intel Saffron, which uses human-like reasoning to find hidden patterns in healthcare data.
Intel is pioneering new technologies that are designed from the ground up for AI, and we’ll talk about new innovation that is purpose-built for the most intensive deep-learning training and inference. This will ultimately help diagnose disease earlier, discover new drugs faster, and tailor precision medicine treatments more effectively.
We’re showcasing more real-world examples of analytics and AI in medicine, including HeartVista, which uses AI to automate the MRI scanning process. This enables fast and accurate cardiac MRI without operator intervention. HeartVista chose an Intel Xeon platform to run the AI for three reasons:
- The need for heterogeneous processing (MRI scanner control, image reconstruction, deep-learning inference, and image displays all must run in parallel)
- The flexibility and rapid deployment required for more than 30 different MRI applications that automatically scale each platform to the Intel Xeon resources
- The use of Intel’s MKL toolbox, both for its MRI-specific reconstructions and via TensorFlow+MKL
Another example is EchoPixel, which uses AI to help accelerate preoperative planning and detect disease faster using medical images. It then brings those to doctors using 3D visualization technology.
Yet another, AccuHealth, uses smart Intel-based monitoring devices that connect to the AccuHealth* Virtual Hospital. Powerful Xeon processors apply data mining and predictive modeling using AI to identify potential trends and concerns enabling AccuHealth virtual hospital staff to proactively contact patients and caregivers to mitigate problems before they escalate and become acute and costly situations. This translates into fewer emergency treatments and admittances means hospitals and clinics free resources to focus on higher value services.
We’re also working with Medical Informatics in using AI to turn patient-generated data into actionable information with Sickbay, its FDA-cleared Clinical Intelligence Platform enabling data-driven medicine. By continuously capturing patient data from any medical device or system, that data is made actionable in web-based clinical apps so care teams can make better, faster decisions and save lives.
And this is only the beginning of the help AI is already providing to healthcare. The future is limitless. To learn more about how Intel is enabling the healthcare transformation, visit our healthcare portal page. You can also keep up with the latest technology trends in health and life sciences on the IT Peer Network.