Exponential Medicine 2019 - Day 1 - Cool Stuff Summary
Day 1 - Monday November 4, 2019
Cool Stuff Summary: this collection of slides and ideas is certainly not a comprehensive summary, but some of the cool ideas I wanted to remember.
The Day 1 live stream is available on YouTube. (I wasn’t able to attend this year’s Exponential Medicine in person.)
Let me know if anything looks incorrect. I’ve tried to summarize the presentations accurately, but haven’t had chance to externally verify what presenters claimed to be true.
[My comments are in italics in brackets]
Trends out of Day 1
Clearly I’m quite biased based, but healthcare system design, when considered at a global context, is clearly emerging as a theme.
1 - Global: It is already quite obvious within the first day of presentations that there is a much stronger focus on healthcare as delivered in a global context, as opposed to what I’ve seen in the past which was much more USA centred. This is great to see, because there are far more people in the world outside of North America who can benefit from breakthroughs in exponential technologies in medicine.
2 - Building systems: the ideas of healthcare ‘systems’ is entering the conversation far more than the past. Before there as more of an emphasis on isolated tools and apps with one-off solutions. This is a step in the right direction to addressing the challenges of health.
The Morning Show
Shawna Butler RN MBA: Nurse Economist, partnerships at Exponential Medicine. (@CleverThought)
Jessica Damassa: Executive Producer & Host WTF Health (@jessdamassa)
Overview of the conference
Day 1: What are the tools?
Day 2: How are the tools applied?
Day 3: How are the tools applied to specific areas?
Day 4: How do you integrate this in your own community?
Session 1
Welcome + Reimagining The Future Of Medicine
Will Weisman, Executive Director, Summits. Singularity University (@wweisman)
Daniel Kraft MD Founder & Chair, Exponential Medicine. Faculty Chair for Medicine, Singularity University (@daniel_kraft)
We wait in the waiting room for hours. Whether we are in San Francisco, or Calcutta.
How do we break down our medical silos?
There is a three phase shift in healthcare: Sickcare -> Healthcare -> Health
Intermittent episodic data results in Reactive Care
Continuous data, hopes to lead to Proactive care
Moving the needle: Reactive -> Proactive -> Preventive -> Predictive
We think medicine today is precise. But in fact it is relatively imprecise. Such as the effectiveness of standard drugs among the population.
“Not just adding years to life, but life to years.”
“Human Uber,” developed in Japan, provides a way to attend events remotely using another person’s body. “It’s surprisingly natural” says its inventor, Jin Rekimoto of Sony.
[This is kind of weird. Will it be more or less weird when robots are the ‘Human Uber’?]
My first sighting of an Autonomous Auto Following Suitcase (AAFS) ... though it seemed like it had too much to drink in the airport lounge... DAFFS. :-) #traveltech pic.twitter.com/OGRAnIWizN
— Daniel Kraft, MD (@daniel_kraft) October 11, 2019
FDA Clears Bio Beat’s wearable cuffless blood pressure monitor
New terms [I’ve never heard these before]
Breathables
Sweatables
Underwearables
Shakeables - for parkinson’s
Pampers now providing sensors in diapers.
Invisibles: can track heart rate and respiratory rate via camera. Applications both in the hospital, or at home
[This is one of my favorite stories of the year.]
Guy fell off bike, was knocked unconscious, and his Apple watch called 911
https://9to5mac.com/2019/09/21/apple-watch-fall-detection-mountain-biker/
AI will help ‘upscale’ your doctor, nurse, or clinician. (see quote on slide below)
[This is a recurring theme of my blog. Alternative terms for this include ‘task-shifting’ or ‘task-sharing’ - being able to move work that was once exclusively the domain of doctors, to be safely performed by non doctor clinicians. This is a fundamental part of how healthcare is dramatically changing]
Session 2
Exponentials Essentials & Updates
Tools to build the future. An update on how rapidly developing technologies are advancing, what is cutting edge today and implications for the future.
2.1 Artificial intelligence & Machine Learning
Neil Jacobstein, Faculty Chair for AI & Robotics, Singularity University
Very powerful computing power, is getting very small.
“You don’t want the events of exponential technologies to be invisible to you”
A number of websites now are offering machine learning courses.
AI can be better the doctors
e.g. Triage - able to outperform dermatologists in recognition of some skin disease
AI can do new from traditional medicine
E.g. Google’s Verily life sciences: able to predict cardiovascular risk factors from retinal fundus photographs; as well as predict the age of the fundus, gender, and hemoglobin A1C (a marker of diabetes).
CBinsights 2019 AI insights infographic.
Problem: 87% of AI projects never make it into production
Therefore, make it into production following a framework such as this:
Limitations of AI system today
CorPath GRX, uses robotics to assist in position of cardiac stent
2.2 Robotics, Humans & Hybridization: The New Surgical Team
Catherine Mohr, M.D. President, Intuitive Foundation
100 year ago, there was trend towards being able to spend more money, and have higher life expectancy.
Now, the life expectancy line has flattened out - because of innovative technology [and new methods of care delivery, and task sharing]
Access to life saving surgery is critical, as a global health crisis. Some numbers:
Countries that have 95% vaccination rates, may still have 5% access to surgery rates.
88% of the next billion entrants into the middle class will be in Asia.
[I love graphs that show world population trends over next 50 and 100 years. It really puts things into perspective for me.]
We’ve stalled increasing life expectancy since 1990s. See how the graph is relatively flat (compared to before) in terms of money spent vs life expectancy.
Why?
Partially because many of the earlier advancements (vacines, and antibiotics) are “Fast infrastructure technologies” - complicated to make, but simple to deploy.
The innovations of the last 25 years, are “Slow infrastructure technologies”, and need more workforce training.
You can build a hospital in a year. But it will take you ten years to train the people to work in that hospital
More examples of fast vs slow infrastructure technologies
Cell phones - fast infrastructure
Physician training - slow infrastructure
NOTE: the life expectancy spending curve, is exponential axis - meaning - we are spending significantly more money without improved health outcomes.
THESIS:
Can we build AI based coaches for surgery, to assist with robotic surgery and tele-mentoring.
Tele-surgery does not solve the problem. Because it keeps the expertise concentrated.
We need tele-mentoring, to be able to train more healthcare workers to safely be able to perform surgery.
How will this change mortality? Lung cancer survival is strongly associated with detection and early surgery. If we could find these earlier at Stage 1.
Catheter robot for lung nodule biopsy, could enable earlier diagnosis.
Can bring additional image augmentation that the human eye cant do. Such as Infrared tagging of anatomy (e.g bile ducts, ureters)
Can overlay 3D images into the surgery.
The foundation’s long term goal: to build a global open-source database of surgical videos, that is regionally appropriate best practices.
[I really enjoyed this talk. A lot.]
2.3 Big Data & Deep Learning
Michael Gillam, MD. CEO & Founder HealthLab (@gillam)
[This was a very interesting presentation. It is hard to tell how much of it is truly working in a reliable peer-reviewed manner, and how much is hype. But the idea is fascinating if successful, but also it is clear to see why many current research publications are of low value]
In short: we used to build cars one at a time, custom pieces of work. We then moved towards automation, which improved throughput.
Michael Gillam makes the same analogy to the process of writing articles. We currently craft these pieces one at a time. This takes a long time. And once published, is already out of date.
THESIS: what if articles could write themselves? Meaning, there is a ‘data lake’ of data that forms the foundation. This is fed into a series of algorithms that study the data. Then interpretations of this data are presented in ‘living articles’ to clinicians. The article always remains uptodate. The data can always be contextual (because it is based on a local data lake).
He presents work where he found 72 areas a human had to manipulate data in order to write an article, and generated a unique algorithm to handle each of these. This could then self generated articles.
Other:
Fastest growing software company in history: UiPath
Google self driving car stopped, for a baby duckling, that was being chased by a woman in an electric wheelchair. Daniel Kraft: ‘we need a LIDAR for healthcare’
Nike has app that uses AR camera’s in smartphone to scan your feet, and provide perfect shoes. Something like this could be used to track edema.
2.4 Genomics
Rodrigo Martinez, Chief Design Officer, Veritas Genetics @RodrigoATCG
Trends
Ancestry DNA fad staring to plateau
Privacy awareness is on the rise
Genotyping is not sequencing
[KEY SLIDE]
Current results of genetic sequencing
3-5% of customers find something ‘life threatening’
Cost of something does not mean consumer adoption. Having a $1000 genome, doesn’t mean that people will buy something. Still need to prove valuable insights from it.
Veritas Genetics dropped the price in July for whole genome sequencing from $100 to $600 in July 2019. Their sales doubled.
Different DNA companies: Medical Utility vs Entertainment.
Do you want to be cheaper and more available? Or more expensive and more valuable
Increased value, by proving more actionable interpretation of the data in a way people can use. The sequencing (the bottom part of the chart) becomes a commodity.
Moving into the Social Genome Era
Challenges over next 3-5 years
Predictions for the coming decade
Session 3
Exponentials Essentials & Updates Part II
3.1 Integrating the Microbiome in Medicine
Jack Gilbert PhD , Professor of Pediatrics, UC San Diego @gilbertjacka
The immune system for many years was seen as an army.
A better analogy is to think of it as a gardener, or park warden. Which aims to keep those parts we want to grow.
Microbiome Wide Association Studies: study technique in microbiome research, to identify the association between microbial activity and human health & disease.
Can transplant feces from a depressed mouse, to a non-depressed mouse: a “re-poop-ulation”
Correlation between gut micro bacteria, and the brain prefrontal cortex imaging.
THESIS: we can change the brain chemistry, by recalibrating the gut microbiome balance.
The Amish, who live close to their farms (opposed to the Hutterites that live further way from animals), have genetically the same lineage, but in testing different types of immune systems.
Exposome (what you are exposed to) -> influences your microbiome
BiomeSense - smart toilet that automatically tracks your feces. [Cool to see this going into production. Just a few years ago this was one of those ideas that ‘may happen in the future].
An example of a “or data-dump”
Currently, only one FDA approved microbiome therapy - the c. diff fecal transplants. Certainly a space to watch.
3.2 Blockchain in Healthcare - The Good the Bad and the Ugly
Heather Flannery Board Chair of Blockchain in Healthcare Global
Mariya Filipova, VP of Innovation, Anthem Inc
Health Utility Network Founding Members
Blockchain: lives in the space between hope & hype
3.3 Virtual, Augmented & Extended Reality
Rafael Grossmann MD, Surgeon & Virtualist @ZGJR
Extended Realities [New term has emerged to categorize these concepts]
Augmented Reality
Mixed Reality
Virtual Reality
Ultra-high resolution displays, e.g Varjo
Cool real time tracking using AR from AnimaRes
Healthcare is a human right
Safe & affordable healthcare.
5 billion without access to safe and affordable surgery.
3.4 (im)Patient Innovation
Richard Hanbury, CEO Sana Health @rhanbury
Standard of care for chronic extreme pain hasn’t changed much over 25 years since his spinal injury.
“Electronic medicine” - ability to change electrical patterns in body through implantable devices
Chronic pain is complex. In addition to the direct pain, there is anxiety, depression, sleep disturbances.
THESIS: Is there a way to get your brain to function, as if it has done 20 years of meditation?
There is lots of data showing the ‘meditation state’ of such brains.
Sana, uses pulsed light to achieve a similar meditation state.
How to create the meditation state, from the ‘outside in’. Such as for people who require the benefits of meditation now (e.g. acute pain episode), but don’t have twenty years to learn this.
Session 4
Convergence, Artistry & Adjacencies
As adjacent fields are converging, new and unexpected approaches to solving biomedicine’s grand challenges are emerging. How can novel forms of visualization enable a better understanding of human and environmental health.
4.1 Visualizing Health & Disease
Alexander Tsiaras, Founder & CEO, StoryMD
Very interesting backstory: he trained as a painter and sculptor, and previously worked as Chief of Scientific Visualization in Yale’s Department of Medicine.
His current company, StoryMD aims to tell a person’s personal health story. The aim to to present their personal health data, in a story form.
They use a sophisticated content management system, that is able to take a patient’s personal health data via an HL7 feed, and match it up with their content library (video, articles, images) appropriate to their unique health data. This is then presented in a highly user friendly format for them to consume.
A person’s personal health history in this way can be presented even as a timeline and story for them to follow.
In addition to their personal health history, there is a library of ‘normal’ findings that the patient can read about too.
THESIS: Goal to be able to tell a story about your health. In beautiful visuals. This is the story of you. Not the statistics of you.
People may consume this information in two ways:
Time of need: you need to know this data know. E.g. you have an abnormal test.
Time of want: you may want to learn more about the data. E.g. on your family member
Move from obligatory to inspiration
Obligatory compliance - do it because the expert says so
Inspiration compliance - do it because you get it.
“Nobody’s life has been changed by a pie chart.”
How do we use big data, into a story.
Problem: HL7 doesn't have a code for love or joy. How do we quantify that in the record?
4.2 Convergence of Frontiers in Veterinary & Human Medicine.
Eleanor M. Green DVM, Dean Texas A&M University College of Veterinary Medicine & Biomedical Sciences
Starter families of millenials are choosing to have pets over children. This is called “Furkids” or “Pet Parents”. [Although ‘cute’, I believe this trend is becoming very problematic for society in the long run]
People spend a lot of money on animals.
Lots of benefits animals on depression, PTSD, stress, alzeimers, homelessness, cancer patients, allergies and immunity, etc.
A number of connections between animal health & human health in spread of disease.
4.3 Fantastic Fungi and Visual Healing
Louie Schwartzberg, Film Director & Cinematographer @LouieFilms
Remarkable cinematography. One of the godfathers of time-lapse cinematography.
Was receiving messages from people who viewed his series, and found they helped their health. Helped advance the idea he should work to incorporate this into healthcare.
Collaborating with Jacob’s Medical Centre, to show images of nature in different ecosystems.
4.4 Exponential Entrepreneurship
Naveen Jain, Founder & CEO of Viome
Session 5
Keynotes
5.1 Changing the Questions: From Healthcare to Health
Rebecca Onie & Rocco Perla, Co-founders of The Health Initiative
Talk not available online.
5.2. Redesigning Healthcare
Stacey Chang, Executive Director of the Design Institute for Health, Dell Medical School. @Stacey_Chang
[This was an exceptional talk. My summary here is subpar. Worth listening in full if you are interested in healthcare system design.]
Stacey previously worked at IDEO. Now moved to Austin Texas, where he leads the Design Institute for Health.
Describes need for a healthcare system.
Not much has changed since the early 1900s when the need for health systems was proposed, from the nature of ward designs, clinic and hospital networks.
Two seperate systems: one of prevention, one of reaction
The things that make us ill today, are no longer addressed well by current ‘healthcare systems’.
The Dell Medical School at the University of Texas at Austin is working closely with the local community to build a new system of healthcare. One with deep community involvement. Starting with principles of imaging care from a blank slate with a new funding model (a local tax).
The following slides demonstrate some of the baseline thought that has gone into how this project is being approached.
Parts of their new system currently have a net promotor score is in the 80s (which is remarkable).
200 highest users of healthcare in Austin Texas, use $40 million dollars annually.
Strategies in other countries.
80% of housing in Singapore is publically owned.
Senior care givers, to care for those 10 years older.
Other
Ways to help patient’s visualize the risk of asymptomatic disease
Exponential Medicine 2019 Summaries
Day 1 (this article)
