5 Ways Artificial Intelligence Is Changing Health Care and What It Means for Quality Management

Artificial intelligence is the future of
health care, and the future is here.
Health care data is expected to reach a staggering 2,300 exabytes (one exabyte is equal to one billion gigabytes) in 2020. We all know knowledge is power, but the amount and complexity of data in existence has long since outpaced the human mind’s ability to access and process it. Enter artificial intelligence (AI). Since its inception over half a century ago, AI has experienced an ebb and flow of attention, investment, development and scrutiny but one idea persists: the promise of making human lives significantly better.

Proponents have long touted AI’s limitless capabilities; critics and the popular media often point to worst-case scenarios where, when left unchecked, the machine intelligence created by human beings not only marginalizes us, but actually leads to our demise. Regardless of these conflicting views and the relative uncertainty of where AI will lead in the future, the technology is already all around us. Increasingly common in business operations and even in daily life (think smart personal assistants like Siri, all those assist features in your car, and the news that shows up in your social media feeds), it doesn’t seem to be slowing down one bit. Among the many “next frontiers” of AI, health care stands to undergo the most significant transformation, taking intelligent systems beyond making our lives easier, to actually saving them.

To get the full details, please view your free e-Book.

A Brave New World of Health Care

Modeled after the human brain and nervous system, AI is a collection of algorithms and information-processing techniques that allow computers to analyze very large amounts of data and then use that data to learn, find patterns and relationships, solve problems and provide insights otherwise imperceptible to human understanding.
Some of the sub-areas gaining most traction in AI include deep learning, robotics, natural language processing, computer vision and Internet of Things (IoT), just to name a few. Thanks to cheaper and more reliable hardware, the Internet’s capacity for gathering large amounts of data, and the ready availability of computing power and storage to process the data, AI has moved from sci-fi story line to reality.    
Examples of new and powerful applications of AI seem to hit the headlines every day, with no apparent limit to where the technology can be used. Recently, Stanford University launched the “One Hundred Year Study on Artificial Intelligence,” a long-term investigation of the implications and potential uses of AI. In the first official report from the study released in September 2016, health care is named as one of the most promising fields for AI. While AI experts have had their sights set on this domain for decades, recent advances and real-world use cases have made it a reality. Here are just five examples of life-changing and life-saving applications of AI in health care.

1.       Cognitive Health Care
In 2011, IBM’s Watson AI system accomplished an impressive feat and caught the world’s attention when it handily defeated two champion opponents on the “Jeopardy!” game show. Fast forward five years and Watson’s powerful technology is in high demand by big industries such as finance, retail, research and, you guessed it, health care. What IBM Watson Health can do that humans can’t is access and make sense of vast amounts of data in very little time. Referred to by IBM as cognitive health care, Watson is able to make sense of vast amounts of medical history and genetic data and provide insights that can be used to diagnose patients, determine the appropriate course of treatment and monitor that treatment. In fact, industry analyst IDC predicts that 30 percent of health care providers will use cognitive analytics with patient data as soon as 2018.

2.       Drug Development
With its ability to parse huge amounts of data and find patterns in ways that humans cannot, AI technology can be an invaluable supplement to the clinical data required to bring a drug product to market. Recently, Johnson & Johnson group Janssen licensed some of its clinical-stage drug candidates to the British AI company Benevolent AI, a tech company that believes its proprietary technology will be able to accelerate the development cycle of promising drug candidates by exploring and providing a rich source of clinical data. “The agreement […] marks a very exciting time for the role of artificial intelligence to benefit scientific discovery and humanity,” says Jackie Hunter, board director of BenevolentAI, in the article.

3.       Digital Imaging
MRIs, cat scans, x-rays and all other diagnostic scans produce data in the form of digital images. These images can hold the key to a patient’s condition and ultimate survival, but they must be expertly read and analyzed by doctors, and used to create the most effective treatment plan.
In an August 2016 announcement, Google DeepMind, a U.K.-based AI company, said it will partner with a cancer treatment provider under the National Health Service (NHS) Foundation Trust, where radiotherapy clinicians from both entities will examine the potential of AI to reduce the time needed to plan the treatment for certain cancers of the head and neck. Because of the delicate and intricate nature of these areas, treatment plans must be mapped out with extreme care, a time-intensive process called “segmentation.” DeepMind believes machine learning can help make this significantly more efficient.
“Clinicians will remain responsible for deciding radiotherapy treatment plans,” a press release states, “but it is hoped that the segmentation process could be reduced from up to four hours to around an hour,” freeing up time for things like patient care, education and research. Eventually, researchers aim to develop a radiotherapy-algorithm that can be applied to other areas of the body. 

4.       Diagnosis and Treatment
Beyond simply retrofitting the AI technology that already exists, developments are also underway to fine tune it to meet niche health care demands. GE Healthcare and the University of California San Francisco are partnering tocreate deep-learning algorithms that will help doctors diagnose and treat patients more accurately and effectively. Best of all, these formulas will be stored in the GE Health Cloud and on “smart” GE imaging machines, which makes the groundbreaking research available to clinicians around the world. According to the MPMN Medtech Pulse article, the ultimate result of these algorithms can be used to predict patient trajectories, automate the triage of routine care, improve process efficiency and enable the development of more personalized therapies.

5.       Detection and Prevention
Prevention is the best medicine and in this vein, wearable medical devices equipped with AI to detect and predict potentially life-threatening events such as blood-glucose spikes and heart attacks are under development.

Also, smartphone apps armed with machine learning, natural language processing and facial recognition can help patients with long-term treatment of chronic illnesses by monitoring medication intake and inquiring about symptoms. Chatbots can even diagnose illnesses using medical data modelsthat link the probabilities between symptoms and conditions based on variables like age, gender, location and time of year. By interacting directly with users to gather information, and drawing from vast libraries of medical information, these systems can provide informal diagnoses, suggest treatments and connect users with specialists, all in a matter of seconds.

The aim of these wearable devices, digital health coaches and medical chatbots is to keep people healthier and help them to better determine when professional medical attention is needed. With that said, none of these AI-powered advances intend to replace doctors or physical clinics, but rather serve both patients and practitioners in the treatment of ailments, whether minor or serious.

New Technology, New Regulations

There is no question that AI offers great promise for a better, healthier future for humankind. But with this potential comes concern about how to maintain control of an intelligence that exceeds our own, echoed among some of the brightest tech minds from Bill Gates to Elon Musk to Stephen Hawking. While the Stanford report concludes that AI is unlikely to ever become capable of the world domination so eagerly perpetuated by pop culture, measured policymaking will be critical to its ultimate viability and public acceptance.
Currently, AI in health care finds itself in a regulatory gray area. The FDA has been slow to address AI in its regulations, primarily out of a need for greater understanding of the technology along with the obvious patient privacy implications it introduces, but also to limit unnecessary interference with advances in the field. While you won’t see terms like “artificial intelligence” or “machine learning” explicitly referenced in current regulations, the FDA has recently issued guidance on three topics which will impact advances in AI.

AI in Quality
So what does all of this mean for the quality management field? AI provides plenty of opportunity for improved efficiency and increased revenue, but only if the technology is used proactively – by humans. Quality professionals will need to expand their skill set to include more technical knowledge in order to take full advantage of these new opportunities. An understanding of areas such as embedded software, data science, analytics, big data, IoT and systems integration will become critical, as they will ultimately change the way in which quality management principles are applied in practice.
Transitioning to an enterprise quality management system (EQMS) is the first step. Some EQMS solutions offer integrated reporting or static analytics, which can be useful tools to capture and analyze a snapshot of data from the past. But imagine a system that could identify the root causes of deviations, or inefficiencies in your workflow, and orchestrate the changes across your entire production system to address them – all within minutes; a system that could predict future outcomes based on historical data. As one of the first in the EQMS space, MasterControl is already leveraging predictive analytics, a form of AI that uses machine learning to assess future performance, in its new solutions.
As the Stanford group states in its report, the success of AI applications is measured by the value they create for human lives. The data that can help us become healthier as people and more effective as organizations already exists. Making sense of it with the help of AI is the way of the future, and it would appear the future is here.

In what ways do you think AI will change the future of quality? Please share your thoughts below!
Beth Pedersen is a marketing communications specialist at the MasterControl headquarters in Salt Lake City, Utah. Her technical and marketing writing experience in the enterprise software space includes work for Microsoft, Novell, NetIQ, SUSE and Attachmate. She has a bachelor’s degree in life sciences communication from the University of Wisconsin-Madison and a master’s degree in digital design and communication from the IT University of Copenhagen.