CONNECTED HEALTH CONFERENCE • SAVE THE DATE • October 16-18, 2019 • Boston, MA
Q: How does technology improve access to care, increase quality of care and reduce costs?
Andrew: When it comes to improving access, reducing costs and improving quality in healthcare delivery, technology is a critical enabler. In the U.S., technology has driven some fundamental shifts in the healthcare sector through digital transformation. We now have a very scalable, rich platform of data to help us re-architect how healthcare will be delivered in the future. Today, many activities that happen in a hospital are recorded digitally. Digital transformation 1.0 was about moving from paper to digital. Digital transformation 2.0 is focused around how to apply the data that has now become readily accessible.
Q: What are the key trends influencing this second wave of digital transformation?
Andrew: At a macro level, there's a big shift happening in how healthcare uses data -- from the traditional use of data for business intelligence or descriptive analytics and more predictive analytics. Today’s focus is on using data not only to evaluate how we've performed in the past, but to look at how we can make better real-time decisions and improve outcomes in the future.
The confluence of larger and more diverse data sets along with the desire to analyze data in real-time has fueled growth in the use of artificial intelligence (AI). AI is a blanket term that encompasses a number of technologies used to analyze different datasets. The AI bucket includes machine learning, neural networks, computer vision, natural language processing and cognitive systems.
As big data gets bigger, the traditional analytics tools for certain types of problems start to break down. All of these techniques under the AI umbrella represent tools that can be used to solve different data problems. Classic machine learning, for example, is good at handling structured data. Deep learning is useful for analyzing unstructured data, such as medical images. An algorithm can be used to analyze images, for example, to help diagnose cancer, tumors, or other anomalies that might be of interest to a radiologist. AI is also a key to unlocking image data. Computer vision technology at scale is evolving to improve the utilization of data from video cameras located in hospitals and other healthcare settings.
Q: How is computer vision impacting operating rooms and other channels for delivering care?
Andrew: Today, many activities that happen in a variety of care settings can be or is captured on video. Computer vision, a tool for capturing and analyzing real-time streams of video, is helping support the development of more distributed care, outside of hospitals and clinics.
In hospitals, computer vision is being implemented for security. For example, looking at the flow of patients, and identifying if someone has left their room or is moving around in the hospital when they shouldn't be. It also can be used for patient monitoring, such as fall detection. Falls both in and out of hospital settings are a major cost to the healthcare system. A number of companies are developing alert systems using computer vision to monitor patients at risk of having a fall.
Wearable technologies also promise to help move healthcare out into the community. We also will see more video captured in real-time during clinical visits in the field. That will open up another channel of rich, new video data assets.
Q: How is AI re-shaping the healthcare sector?
Andrew: Today, healthcare organizations are customizing AI strategies to solve problems based on the unique types of data they are using. From a healthcare provider perspective, there are currently three primary use cases for how AI gets deployed. One is an operational use case, which help hospitals reduce waste and improve efficiency. Second, is a financial use case in which AI is helping optimize processes, such as billing and accounts receivable. Finally, there are clinical use cases in which AI is being used to improve patient care.
While the clinical use cases get a lot of attention, data scientists in healthcare will tell you that they start with operational and financial use cases. For most organizations, starting with the financial and operational use cases gives them the ability to deploy AI in a lower acuity environment with clear, near-term ROI before transitioning into direct patient care.
AI also is being used to streamline patient experiences. We’ll increasingly see it used in new types of distributed and remote care settings. AI is facilitating access to big sets of population health data, which can be used to provide more personalized care to patients anywhere in the healthcare ecosystem.
Q: What is Intel is doing from an AI and computer vision perspective to make all these new sources of data actionable?
Andrew: The crux of all of this amazing technology is the output of data and information. Today, we're generating gigabytes of data per person every day. Intel technologies reduce the barriers for capturing, analyzing and securing this data. This starts from bringing down the cost of computing, storage, and networking. This enables the core infrastructure to collect, to move, and to process that data. It also makes it economically feasible to do more analysis, and to combine very large datasets into single datasets that can be used to unlock new insights. The business case makes sense. With cheaper computing and cheaper storage, you can start to build a more favorable ROI model for making investments in the infrastructure to support that.
Intel is facilitating the development of product roadmaps that use AI for the analysis of large datasets by healthcare organizations. As a company, we recognize the value in data, and we're developing the infrastructure to make it possible for every organization to be able to collect, analyze, and extract insights from the data they generate.
Andrew Bartley is Senior Solutions Architect in Intel's Health and Life Sciences Organization. Intel is the Partnering Sponsor of the 2018 Connected Health Conference.