|
Getting your Trinity Audio player ready...
|

“I’ve been doing different things in artificial intelligence for about 13 years. In general, I see AI as a tool and not a solution. The number of tools in this category called AI is ever-expanding which is both exciting and overwhelming.
Healthcare is so fundamental to everything, and everything is fundamental to healthcare. The ability to pay the bills, to buy decent food, to pay the rent, to have a home – that’s going to dramatically affect your health and well-being.
At a national scale, we spend $5 trillion a year on healthcare. That’s a number that’s difficult to even fathom. That equates to 20% of our GDP, and it’s only going up. That’s one of three massive circles in a challenging Venn diagram. Another is that we have a population with growing healthcare needs. And the third Venn diagram circle in this unhappy triad is that we don’t have enough clinicians to take care of all those patients if we could afford it, and we can’t.
“We have the significant potential to live happier and healthier lives.”
Dr. Thomas Osborne
So the old paradigm – the traditional path that we’ve tried to solve these problems – hasn’t worked. The way I see it, the only way that we can solve these types of challenges is to do it differently, to take a different approach. Thoughtfully implemented technology has the potential to address all those massive challenges. If we can do that, then instead of going into a vicious cycle, we can spin that around and turn it into a virtuous cycle.
There’s probably at least three general areas [of using AI in medicine]. There’s the back-end administrative operations, there’s research and development [and] discovery, and there’s also direct patient care.
Operations and efficiency… that is probably the lowest risk and quickest value use of these tools. We spend a lot of money in healthcare and an estimated 25% of that is waste. This may come in many forms that are solvable with AI, such as ordering too many medications that then expire. However, intelligent decision support systems that learn the seasonal variations, epidemiology and perhaps even social media trends can assist data informed decisions about what and when to purchase supplies proactively. There’s not only waste, but also inefficiencies. A large proportion of [physicians] said that they spend 20 hours a week on paperwork. That’s a poor utilization of a rare and expensive resource. This repetitive paperwork, and data entry is well-suited for automation with AI empowered tools.
On discovery… Pathology and physiology is extremely complex and difficult to fully understand with traditional tools. However, AI-empowered analysis of complex data can and has unlocked new insights leading to better understanding of pathology, new drug discovery and care management. The frontier of possibilities is also expanding in many ways. For example, imagine a patient has a challenging rare condition and it is not clear which treatment of the approved paths is optimal for them. In the near future, we may have intelligent health record systems that can securely compare people that have faced similar challenges in the past, who have similar backgrounds, and then identify which treatment path was most successful. These data-informed personalized care pathways can lead to better outcomes at a lower cost.
On that same note, AI-empowered research can leverage other tools. For example, we can now visualize in great detail what molecules look like, which then informs us how they work. Since our body runs on the performance of these molecules, understanding them unlocks pathways to accelerate new drug discovery as opposed to the traditional and longer way of finding new medications by luck, trial and error.
Moving from research into clinical care, imagine you’re [a patient] waiting to be seen at a busy doctors office. While you are waiting, a staff member takes you to a private room and has you look at a camera that takes a few pictures of your eye. The pictures of your retina can then quickly and noninvasively identify risks about cardiovascular disease and other conditions earlier than they would have been, leading to better and cheaper care. In addition, the quality of your voice may be a biomarker that can help improve the detection of an important condition when it is easy to treat. These types of advancement are right around the corner, and can enhance care to many more people, especially to those who may not have access to specialists.
There are many of us working to ensure that any medical equipment, AI or otherwise, is developed and used safely and responsibly. It is important that these tools are rigorously tested and evaluated so that they are optimized for the most benefit in the right situation. However, it is also important to make sure that life-enhancing advancements are not just available to some. For example, how can we justify not using tools that are so much better than the standard of care? What if we had something that could prevent or cure a horrible disease… and you didn’t use that. Would that be unethical? I think it would be.
This is the most important and exciting time in history. We have the significant potential to live happier and healthier lives. However, we need to also recognize that, at this time of unprecedented rapid change, it can feel very overwhelming. Together, we have the ability to leverage our strategic advantage of diversity in technology, thought, and experience to help direct these advancements for shared prosperity.
For example, how can we offload the drudgery of repetitive work to allow more time for patients, business strategy, or creativity and innovation? How can we enable more people to invent, start businesses, and explore their dreams? How can we do this while also having more time for family and friends, to live more fulfilling lives, and to be more human?”
Dr. Thomas Osborne is a physician and researcher. He is the chief medical officer for Microsoft Federal and an adjunct professor in the department of neurosurgery at Stanford University.
This profile is one of nineteen featured in the “Humans in an age of AI” portrait series. See more profiles here.



