Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:03):
Welcome to the House of surgery,
a podcast brought to youby the American College of Surgeons.
In this series, you'll hear fromand about surgeons in all specialties,
in all practice configurations,and in all locations,
their success stories, advice, challengesthey've overcome,
and words of inspirationas they serve their patients with quality,
(00:26):
integrity and professionalismand strive to heal all with skill
and trust.
The opinions
expressed in this podcastare those of the participants
and not necessarilythose of the American College of Surgeons.
Enjoy the program.
(00:47):
As artificial intelligence applicationsand offerings continue
to develop and evolve, surgeons are facingnew questions about how to ethically
and responsibly utilize this technologyin both academic and clinical practice.
I'm Dr. Ali Mohamed,neurosurgery resident at Mayo Clinic in Florida.
kneed of the ArtificialIntelligence Committee of the Council
of State Neurosurgical Societies, premiersocioeconomic body of organized neurosurgery.
(01:09):
Virtually is Dr. Brandon Lucke-Woldneuro endovascular fellow
at the University of Floridaand vice chairman of the Workforce
Committee of the Council of StateNeurosurgical Societies.
This episode of the House of Surgeryfrom the American College of Surgeons
explores the topic of AI and surgerywith Dr. Daniel Donoho, pediatric
neurosurgeon at Children'sNational Hospital
and founder of the Surgical Data ScienceCollective,
(01:30):
a nonprofit research organizationdedicated to improving surgical outcomes
by leveraging the powerof artificial intelligence.
He's also a policy advocateand thought leader in data science
and surgical AI,both within neurosurgery and beyond.
This episode will focus onwhat responsible and ethical AI adoption
looks like and means for surgeons,particularly when it comes to recent
(01:51):
controversial topics such as the use of AIin prior authorization.
Dr. Donohothank you for joining us today.
Thanks so much, Dr. Mohamed.
It's really a pleasureto be with the both of you today.
Coming to you from the AANS annualmeeting, here in San Antonio, Texas.
Excellent.
So I think just to kind of start off,maybe you can tell us a little bit
(02:11):
about your experience, in data scienceand artificial intelligence,
how you got involved.
And, where do you thinkwe are now as a discipline?
It's been a
pleasure to watch so many peoplebecome involved in this effort
along with us.
For me personally, this effort started
as a way to understandhow we can improve surgery.
(02:31):
I was working in a projectin surgical simulation
where we were trying to learnhow to better manage
a kind of vascular injurythat we encounter in the operating room.
And it became clear that we neededa new set of tools
to understand a new type of data thatwe were generating, which were images.
At the same time,we were also working on many classical
(02:51):
big data projectsor national databases of the like that
many podcastlisteners will be familiar with.
And as we continue to work in this area,the intersection of these two worlds
of computer vision and machinelearning and deep learning
and artificial intelligence,really became possible.
And that's one of the reasonswhy I'm so excited about this
(03:11):
technology is its ability to uniteso many previously disparate types of data
and make actionable recommendations,provide useful
outputs to hospitals, to surgeons, and to many other surgical stakeholders.
So that was a great, introduction.
We really are interestedin hearing a little bit of your thoughts
(03:33):
on what role the surgeon playsin interacting with data science and
how surgeons can use their leverageto impact change.
I think surgeons have unique leveragein the perioperative setting.
That's both because we understandthe problem the best.
So Professor Rich Byrne who is the Chair, at Mayo Florida
(03:57):
has, likes to say that, every surgeonshould think of themselves as an engineer.
And we're all facing engineering problems.
We have unique ability
because we understand these problemsto generate solutions to them.
The same thing is true in the dataand AI world.
We also have unique influencewithin the hospital setting.
We have a unique position within the valuecreation chain in medicine.
(04:21):
And those thingsallow us to bring a perspective
that is extremely valued by institutions,and we can serve as trusted
advocates and change makerswithin a hospital and beyond it.
One of the things that the three of us allwork on is how to create
optimal policy and guidancearound the implementation of technology,
(04:44):
around the future workforcethat we will all be a part of.
And we are facing a technologythat, for the first time,
may actually fundamentallychange the workforce in healthcare.
It may change the workforce around us.
It may change how we work.
It may change the tools and settingsthat we work with and work in,
(05:05):
and it may even change the jobsthat we as surgeons do.
So I think that any surgeon in 2026
needs to see themselvesas part of an arc of a dynamic career
that is different than it wasten years ago.
Even if some of those differencescan't be seen yet,
they will become more evidentover the next years to come.
(05:30):
So we've talked a little bitabout advocacy in terms of, you know,
policy and responsible useand a little bit about the surgeon's role.
But we know that these
AI applications and implementationscome from industry side as well.
So what are some of the challengesyou kind of foresee with kind of this,
a realitythat AI is not just coming from us?
And what role do we have to, you know, navigate
(05:52):
these new technologieswhile also advocating for our patients?
So a little while ago,someone made a seemingly
outlandish statementthat there would be a single employee
billion dollar annual recurring revenue
company created in the age of AI.
And this would happenat some point in the future.
(06:13):
Recently, the New York Times ranan article about someone
who ostensibly did that in health care.
Now, there are a lot of problemswith that particular,
you know, that were revealedand subsequently to that story.
However, I think it servesas an interesting counterpoint.
In the talk I gave to the CSNS,we talked about the capital expenditure
(06:36):
that's going into artificialintelligence is approximately the size
of the largest health medical devicestrategic company per quarter.
So we use the analogy of
saying that, the Magnificent Seven,soon to be the Magnificent Ten perhaps,
AI companies are spending
one medtronic'sworth of capital expenditure per quarter.
(06:58):
This is a area
that is, contradictory, right?
On the one hand, right.
You have individuals that can actuallyservice companies themselves.
And on the other hand,you have companies operating at a scale
that we've never really seen or understoodbefore in our era, in our field.
(07:20):
So I think that it's importantto embrace the possibility
and opportunity that new tools bring,
especially for our younger audience,
but also for our establishedsurgical audience.
So, in past technological revolutions,the onramp has been frustrating.
So it has been more natural to say that,“Oh, this is a product
(07:43):
that's going to be adopted by someonewho's in their teens or early 20s.”
I think about, for example, one of my colleagues,
who probably spends on the orderof several million tokens
per day,running in the genetic coding swarm.
And he is, let's just say not
(08:03):
in his first few years of practice, but,
he has embraced very thoroughly
this AI revolutionand is really using it to change his field
to develop new tools for himselfand for his colleagues.
And to really sort of liberate himselffrom what is otherwise
been a very dependent position on industryand even on other scientists.
(08:26):
So, I think that this is a technologythat makes itself easier
and easier to usewith every successive iteration,
and has some unique opportunitiesthat may make us less dependent
than we would have been on industryto put, let's say, a pedicle screw
or an aneurysm clip out onthe market to distribute.
(08:47):
Ah, excellent points.
in terms, some of those like changesespecially like,
are coming, on the horizon.
A lot of the interaction is based
on some collaborationsoutside of our own discipline as well.
Can you talk a little bitabout some of those collaborations?
(09:10):
How those are developed and then what
you envision in terms of, needs, both in terms of an academic
and continued partnership, with some of the industry players as well?
I think that as surgeons,
right?
And, and for this audienceand the American College of Surgery,
the House of SurgeryI think it's critical for us
(09:34):
to work outside of our silos,within our own domains.
It doesn't make much sense.
So if you think about another industryby analogy,
let's say the restaurant industry,you don't necessarily fundamentally
operate a Thai food restaurantdifferently to an Italian food restaurant.
(09:54):
They're not separate subcategories.
Yet in surgery,we like to slice down these differences.
One of the reasons that I'm passionateabout working with the College,
and I was privileged to participatein the Clinical Congress
last year, and will do so again.
So come see us inWashington, DC in September.
Is that we need to break outof our small disciplines.
(10:18):
Neurosurgeryis a few thousand of surgeons.
Even a general surgery viewed as a fieldis a small fraction
of all of the proceduralists.
All of the image guided intervention listsand all of the physicians,
within the United Statesand then within the world.
There is a lot of innovationthat is happening all over the world.
(10:39):
And one of the critical factors
that we work towards on a dailybasis is how can we bring disparate
methods and disparate people
together to learn and to improve?
In our organization,in the Surgical Data Science Collective,
this happens not only with our projects,where we might bring a team from Kansas
(11:04):
together with a team from Kenya,
but it also happens in the kindof technical methods that we use.
Where we might bring in somebodywho's worked in autonomous vehicles
with somebody who's worked in agricultureor somebody who's worked in big tech
together with a surgeonwho practices in the Midwest.
(11:24):
And together, that team becomes stronger
than its individual constituent parts.
How we do that,I think, is by working with organizations
that inherently understandthe need to transcend boundaries.
I think that's where surgical societiescan be extremely useful,
particularlythose that work across disciplines.
(11:44):
How this relates to industry,which was the last part of your question.
I think that when we work together
and when we have vehicleswhere we can individually
and collectively benefitfrom our collaborations,
we present a very different imageand structure to our industry partners,
rather than being viewed potentiallyas single individual KOLs,
(12:07):
with all of our foiblesand knowledge and faults and good parts.
We actually can form cohesive entitiesthat can provide them with aggregated
knowledge and experience that can betransformative for their products
and advance their industries.
So I think some of those structuresare also important to consider.
I would be remiss if this conversationended, and, we didn't talk a little bit
(12:27):
about some of the controversial topicssurrounding AI.
Particularly, the use of AI and prior authorization.
What are your kind of thoughtsregarding this phenomenon,
how do you think we move forwardin a responsible and ethical manner?
I think that as physiciansand as patient advocates,
we have some unique constraintson our behavior that perhaps other actors
(12:50):
in this space may feel less strongly.
But we also have some unique strengths.
I think it's importantthat we embrace our strengths
and use them to our fullest advantage.
I think that technologycan be very patient centered.
Our unique positionas advocates for our patients
allows us to deploy technologies strongly
(13:12):
and in a forwardthinking manner to support our patients
because it's the right thing to do.
When we are taking care of someone
who's losing neurologic function,who has intractable pain, who can't work,
and for a payor to deny that care,
we have some unique leverageto fight back.
We also have unique leverage,perhaps, to change
(13:34):
the kind of conversationwe're having with payors who,
if I'm being honest, I'm not sure thatthey would characterize this as a fight.
I think they wouldcharacterize their behavior
as trying to align funding for health carewith how it should optimally
be allocated, and who better to decidethat than physicians?
(13:54):
I think our computer systems,
our tool using large language models
that are facingtheir first real world uses
in the operational sideof healthcare and document generation,
communication, coordination, scheduling,revenue recovery and billing
those areasthat look most like normal businesses.
We have the ability to use those as well,
(14:17):
and I think we shouldethically and carefully.
I think that we might findand I was talking about this
with, Dr. Timmy Allen, right.
I think we might find thatthere's more opportunity
for us to align our surgical plansand our requests for patient care
with the guidelines of payors,by using tools that understand
(14:40):
those guidelines a little bit better.
Hm-mmm (affirmative).
Those tools could be deployed at or very near the point of patient care,
to make sure that all of us have the bestand easiest time at getting appropriate
care reimbursed quickly and efficiently,which is all of our goal in this space.
So I think we should push forward.
I think we shoulddevelop patient centered,
(15:02):
and care first models
and software that can advance our goals.
And those systems
should be able to be in dialog directlywith the systems of the payors.
Some of whom could be governmentactors or private actors.
And I think
we could develop thosein a variety of different structures.
But as surgeons, we should lead it.
(15:24):
Absolutely.
Yeah, excellent points.
I think part of that comesto a little bit of the ethics
in terms of the decision making.
There's been like an exponential increasein publications, like recently.
And really disseminatingthat information,
(15:44):
in this era of large data science.
Can you speak a little bitin terms of like, guidelines and what
from a surgical standpoint,how we push advocacy for our patients
and in ensuring the appropriate ethics,with this large amount of new data
that is coming outthat is feeding into these algorithms.
(16:07):
It's going to be extremely complicated.
And so, in a complicated area,the answer is always to lean in
and to lead.
That's what we do as surgeonsand that's what we need to do here.
So we're developing systemsthat can find and surface
patterns and relationshipsthat were not conceived of at the time
(16:28):
that other data was collected.
And I think this creates uniqueopportunities and unique challenges.
The best way to deal withthose is always for us to lean in.
There are ethical challenges using AI.
Those are going to be pushed upon us.
We're going to see patients using AI.
We're going to see hospital systemsusing AI,
(16:51):
and we're going to have our own desiresto use these systems
to do a lot of different thingswithin the patient encounter.
We need to understandnot just the attitudes or desires,
but actually how these systems workand how they're going to interface
with our processes of careand where ethical concerns rise up,
what matters most to patients.
(17:11):
And I think we can do it,and I think we should lead it.
Dr. Donoho,
thank you for joining us todayto discuss all of this.
Before we close,
is there anywhereyou can direct the listeners
to learn more about, AI adoption and surgery?
And some of the other topicswe discussed?
Yeah,I think, so a couple of things, right.
From the American College of SurgeonsClinical Congress,
(17:34):
which will be in Washington, DC.
You should attend.
We’ll be there sharing our perspective in a great dynamic session around AI.
We did one last yearthat was well attended and also broadcast
on the internet.
The second is our nonprofit research
organizationhas a website at: www.surgicalvideo.io.
(17:55):
We encourage you to learn more
about what we're doing in surgicalAI and to follow
along with what we're doing onvarious social media properties, as well.
And lastly, to invite collaboration.
That's really the point of all of the work that we're doing.
We work with general surgeons,ObGyn surgeons, ophthalmologists, ENTs,
(18:15):
and many other disciplines,even outside of surgery.
And our missionto advance surgical quality and achieve
better patient outcomes.
So thank you all for this opportunity.
And really,it's been a pleasure to be on the podcast.
Phenomenal. Thank you again. Thank you.
Thank youfor joining us on the House of Surgery®
(18:36):
podcast, brought to youby the American College of Surgeons.
If you like this podcast,please rate it five stars
and let your friends and colleaguesknow about the podcast.
On social media,use the hashtag #HouseofSurgery.
You can learn moreabout the American College of Surgeons,
its members, programs,products, and services at facs.org.