Dr. Anthony C. Chang from the Children’s Hospital of Orange County wants his fellow doctors to think outside the box when it comes to knowledge and expertise. Can an ‘AI Sherlock’ help?
By Sophia Stuart
In the medical community, a clinician’s opinion holds sway, so wariness among professionals about artificial intelligence is not surprising.
But Dr. Anthony C. Chang, practicing Pediatric Cardiologist and Chief Intelligence and Innovation Officer at the Children’s Hospital of Orange County (CHOC) saw the writing on the wall several years ago and has embraced the emerging technology. He even went back to school after several decades as a doctor to study biomedical data science and AI at Stanford School of Medicine.
Today, he’s often referred to as “Dr. AI,” speaks regularly at Singularity University Exponential Medicine conferences, and presides over the Medical Intelligence and Innovation Institute (MI3), the first of its kind inside a hospital, funded by the Sharon Disney Lund Foundation.
We sat down with Dr. Chang at the AIMed symposium in Laguna Niguel, California, in December to talk about his pioneering work and how he wants to build a Sherlock-style AI clinician-computer scientist interface. Here are edited and condensed excerpts from our conversation.
Let’s start with your concept of “AI Sherlock” as a doctor’s partner.
I realized that when we keep gathering together physicians from similar backgrounds and mindsets, we go back and forth with the same issues, ideas and controversies, and whomever has the highest decibel voice wins the argument. I knew there had to be another way to get to the answers than just through evidence-based, as opposed to data-supported, medicine. I thought it would be wonderful if every doctor had an AI-based partner like the data-savvy and objective-thinking Sherlock.
Which is an unusual view in your profession.
[Laughs] Yes, especially 10 to 15 years ago. Because, as doctors, we’d been trained to memorize publications and results, considering those absolute truths from some oracle. I’ve always been something of a radical and didn’t want to take that position necessarily. I saw so much information contained in records, experience gleaned from other doctors, and I knew there had to be a way to organize and interpret all this data to improve medicine, to create a “super opinion” of some kind.
So you took the brave step of going back to school, after decades of seniority in your profession, and attended Stanford.
I love to be out of my comfort zone, because I learn so much there and biomedical data science was starting to become a real program, with the one at Stanford being among the first. In fact, I downloaded my application to Stanford on February 16, 2011.
The night IBM Watson won Jeopardy.
That was a watershed moment—and a wake-up call.
Are you a different sort of doctor now, post-Stanford?
After Stanford, I felt as if I had a more complete brain, using both the analytic/procedural thinking data scientist aspect and my training in medicine. I thought this new knowledge would be additive, but it’s actually geometric, in the way it’s enlarged my brain, giving me a deep insight and perspective into what’s coming next.
Which is why it makes sense you’re both the Medical Director and Chief Intelligence and Innovation Officer at the Medical Intelligence and Innovation Institute, at CHOC today.
Our patients are too valuable not to have the best of both worlds at the top today.
It’s clear many in your profession are coming round to your point of view. Here at AIMed today, the conference hall is full of very senior physicians, eager to learn about medical intelligence, which is unusual, as tech-focused conferences tend to skew much younger.
I’m very happy about that. I feel that less than 5 percent of doctors currently understand the magnitude of AI’s entry into medicine. Meetings like this are great, because we have an even split between clinicians and non-medically trained personnel. There has been too much depletion in other fields, as data science replaced experienced staff, I don’t want that to happen in medicine. We have 20 countries represented here at AIMed, with many more asking if we can bring the next one to their part of the world, and we’ll be going to Asia and Europe next year.
Symposia aside, can you talk in practical terms about the AI-based clinician-computer scientist interface you’re building now?
It’s called LEONARDO, after the Renaissance artist/scientist Leonardo Da Vinci, who, in 1513, was the first person to describe a congenital heart defect, [there’s a paper here, if you want to learn more]. It’s an interface between clinicians with data and computer scientists with analytics to assure a data-to-information continuum and eventually a knowledge-to-intelligence transfer. Drawing on advanced computing, graph databases, computer vision, natural language processing, IoT, neuromorphic computing, genomics and deep learning, LEONARDO combines both the current state of understanding and research, with the seasoned wisdom of sagacious senior physicians, while minimizing diagnostic and therapeutic errors. Within 20 years, I believe, it will become a tireless and omnipresent multilingual supermentor for generations to come, working alongside human cardiologists in AI-based clinical synergy.
The rapid development of AI within all fields is exciting, do you ever want to take off your white coat, permanently, and head north to Silicon Valley?
It would be very tempting to stop practicing medicine and dive into advising and setting up startups in the Bay Area, but I feel a need to see patients every day to fully understand the pain points and realize what we can do for them using all this advanced computer technology.
If you want to learn more, Dr. Chang’s Medical Intelligence and Innovation Institute (MI3) holds regular monthly meetings, open to the public, with invited speakers. They will also be holding an Innovation Beach event on Feb. 21 in Orange, CA. AIMed returns in December 2018.
Disclosure: Sophia Stuart stayed at the Ritz Carlton Laguna Niguel as a guest of AIMed to attend the conference.