March 25, 2025 • 27 mins
Artificial intelligence is no longer the future of veterinary medicine—it's very much the present. In this captivating discussion with guest editors Casey Cazer, Parminder Basran, and Renata Ivanek, we explore the groundbreaking AJVR supplemental issue "From Bark to Bytes: Artificial Intelligence Transforming Veterinary Medicine."
The conversation reveals how AI applications already extend far beyond the clinical notes scribes that many practitioners might be familiar with. Veterinarians are now using AI-assisted stethoscopes to detect bovine respiratory disease, employing machine learning algorithms to predict Lyme disease risk patterns, and leveraging artificial intelligence to fill gaps in antimicrobial resistance surveillance data. Each application demonstrates how this technology can enhance clinical decision-making while accelerating vital research.
Our guests emphasize that successful AI implementation requires multidisciplinary collaboration, quality data, and thoughtful integration. "Garbage in, garbage out" remains a fundamental principle—without standardized, high-quality data, even the most sophisticated AI tools will produce unreliable results. The ethical dimensions of AI in veterinary medicine also take center stage in our discussion, from ensuring data privacy and informed consent to recognizing inherent biases and maintaining the veterinarian's ultimate responsibility for patient care.
For practitioners curious about incorporating AI tools into their workflow, our experts recommend starting with well-researched technologies, implementing them gradually, and evaluating how they affect the veterinarian-client relationship. As this field continues its rapid evolution, staying informed through resources like this supplemental issue becomes increasingly crucial for veterinarians who want to harness AI's potential while navigating its challenges. Join the conversation at the second Symposium for Artificial Intelligence in Veterinary Medicine at Cornell University (May 16-18, 2025) to explore how these technologies can help shape the future of animal healthcare.
INTERESTED IN SUBMITTING YOUR MANUSCRIPT TO JAVMA ® OR AJVR ® ?
JAVMA ® : https://avma.org/JAVMAAuthors
AJVR ® : https://avma.org/AJVRAuthors
FOLLOW US:
JAVMA ® :
Facebook: Journal of the American Veterinary Medical Association - JAVMA | Facebook
Instagram: JAVMA (@avma_javma) • Instagram photos and videos
Twitter: JAVMA (@AVMAJAVMA) / Twitter
AJVR ® :
Facebook: American Journal of Veterinary Research - AJVR | Facebook
Instagram: AJVR (@ajvroa) • Instagram photos and videos
Twitter: AJVR (@AJVROA) / Twitter
JAVMA ® and AJVR ® LinkedIn: https://linkedin.com/company/avma-journals
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Wellness Pet Food (00:01):
Mealtime is more than a bowl of food.
It's the foundation for a longand healthy life.
At Wellness, dry recipes arecrafted with the highest quality
natural ingredients for a tastethat pets love and
scientifically proven throughAFCO feeding trials to ensure
that every complete and balancedmeal supports the five signs of
wellbeing.
With Wellness Pet Foods, petparents can share a life of
(00:25):
well-being together with theirpet.
Sarah Wright (00:31):
This is Veterinary Vertex, a podcast of the AVMA
Journals.
In this episode we chat aboutthe March AJVR AI Supplemental
Issue From Bark to BytesArtificial Intelligence
Transforming Veterinary Medicine, with our guests Casey Cazer,
Parminder Basran and RenataIvanek.
Lisa Fortier (00:50):
Welcome everyone.
I'm Editor-in-Chief LisaFortier, with co-host and
Associate Editor Sarah Wright.
Parminder and Casey welcomeback.
Renata thank you so much forbeing here with us today.
Great to see my Cornellcolleagues.
Renata Ivanek (01:03):
Happy to be here, thank you.
Parminder Basran (01:04):
Yes, absolutely.
Sarah Wright (01:06):
All right, let's dive right in and talk all about
AI.
So, you are the guest editorsfor the AJVR supplemental issue.
From Bark to Bytes ArtificialIntelligence, transforming
Veterinary Medicine.
Casey, what can readers expectto take away from these articles
?
Casey Cazer (01:26):
In this special issue, I think readers will find
papers written by differentleading experts and researchers
who are using AI in veterinarymedicine.
These authors were actuallypresenters and attendees at the
first annual Symposium ofArtificial Intelligence in
Veterinary Medicine that washeld at Cornell University in
April 2024.
So, whether you are unsure whatan LLM is or whether you're
using AI tools every day in yourveterinary practice, I think
(01:49):
there's something new foreveryone to learn in this
special issue.
I would encourage readers tostart with the first two
articles in the issue, whichprovide sort of a gentle
introduction to AI in vet medand also the FDA's regulatory
perspective on AI and machinelearning in veterinary medicine,
and the takeaways that I tookfrom these two articles is that
(02:10):
AI can be very useful right nowor in the future in many aspects
of veterinary medicine,including diagnostics, drug
development, clinical decisionmaking, record keeping and
client communication.
However, there are a lot ofpotential pitfalls, including a
need for education ofveterinarians and clients on
these AI tools, a lack of highquality data to train the
(02:32):
machine learning models, andethical and trustworthiness
challenges.
Lisa Fortier (02:38):
Yeah, it was really fascinating.
I enjoyed reading every one ofthe articles.
I'm not honest, I didn'tunderstand all of them, but I
certainly learned so muchreading them.
Just the breadth of how AI isalready involved in so much and
what the potentials are.
It's really fascinating.
So thank you for the idea toform this supplemental issue.
Casey Cazer (03:00):
Sure, so I hope that if people are interested in
this topic, they want toconsider attending the Symposium
for Artificial Intelligence inVeterinary Medicine 2.0, which
is going to be held at Cornellon May 16th to 18th 2025.
And you can get moreinformation on it at
cornellaivet.
org.
Lisa Fortier (03:20):
Perfect, that's when our forsythia is in full
bloom.
So come on out to Ithaca.
Hey, Parminder, what sparkedyour interest in artificial
intelligence and then with yourcolleagues Casey and Renata,
forming this symposium?
Parminder Basran (03:36):
When I first joined Cornell about almost six
years ago, there had alreadybeen a fair bit of research in
human medicine and artificialintelligence that I was
undertaking working in a humanhospital and when I arrived here
I realized that there wasn'tnearly the same amount of push
(03:57):
or, I guess, projects and theamount of resources that people
are spending in veterinarymedicine and artificial
intelligence.
So that to me reallydemonstrated sort of an unmet
need needed within ourenvironment here in veterinary
(04:18):
medicine.
So, there was a natural sort ofprogression of exploring
artificial intelligencemigrating from human medicine to
veterinary medicine.
And then soon after that, it'samazing how long ago two years
feels from now.
But if we were to go back in atime machine and think about
what life was like two years ago, there were tremendous
(04:46):
questions that we were askingourselves about artificial
intelligence and veterinarymedicine.
You know, things like ChatGPTweren't even on the radar and so
at that time a bunch of us gottogether and said, hey it would
be great if we could create anenvironment where we can bring
together people from differentdisciplines in human medicine
and in epidemiology and incompanion animal medicine and
(05:11):
computing science and engineersand just really talk about what
artificial intelligence means inthis space, and so that was the
germ of the idea is for us toreally just start having a
conversation about this a coupleof two years ago, and what
happened after that was we wereable to secure some funding
(05:34):
through the NIH and the FDA tosponsor the event, brought
together sort of these fourpillars of veterinary medicine
companion animal, populationmedicine, comparative health
basically bringing togetherdifferent components of
artificial intelligence andveterinary medicine in a single
(05:56):
setting.
We also realized at the timethat there weren't a lot of
avenues for our trainees to gettogether and talk about
artificial intelligence inveterinary medicine.
There may be a lot ofspecialized AI conferences, but
there wasn't really anythingthat really captured veterinary
medicine.
So again, trying to see if wecould create a safe space for
(06:21):
people of different disciplinesto get together and talk about
these important things wasreally important.
And so having that funding fromthe FDA and the NIH helped
really push the symposium alongand we were really happy about
that.
And we were very pleased thatwe helped kind of break down
some of the barriers forknowledge and education in that
(06:41):
process as well.
So you know, the more that wecan share knowledge about things
that we don't really understand, the less you know, the less
fearful we are of thattechnology, and so we were
pretty happy that we were ableto do this thing and, based on
our success of that event, it'sreally evolved and become
(07:06):
something of a beast of itself,so they're pretty happy about
that.
Lisa Fortier (07:17):
Yep.
Success comes with someresponsibility, then, doesn't it
?
Yeah, walking around at theposters and reading about the
speakers, how diverse theirbackgrounds were in coming and
their disciplines, and how it'sthe most collaborative field of
all it was fascinating.
(07:38):
Well done to get that sort ofattention as well on your first
one!
Parminder Basran (07:40):
I think that's a great point in the sense that
the thing that we know aboutartificial intelligence and
having successful practicalapplications of artificial
intelligence is that it can'tjust be done in a silo.
A single person can't do, can'tget success, uh by by, you know
, going alone.
So so bringing people withdifferent expertise and
(08:03):
disciplines together was a bigpart of what we wanted to do,
and indeed that really holds upin terms of the literature, in
terms of the successfulness ofAI adoption and implementation
in clinical practices, is thatthose environments where that
have lots of multidisciplinaryteams, people from different
backgrounds, tend to do verywell in deploying AI in their
(08:27):
communities and environments.
Lisa Fortier (08:28):
That's fabulous, Casey.
Back to you for a minute.
Sarah asked you a littleearlier what were some of the
interesting take-home messagesfor folks, but always, at least
for me, there was lots ofsurprises, both at the symposium
and in things I just didn'tknow.
What things surprised you inthe supplemental issue, what did
you learn that was like huh, Ididn't know that?
Casey Cazer (08:49):
Well, I think, actually similar to you and, how
you said, walking around thesymposium and looking at the
posters, you were surprised athow much diversity there was.
I think that in thesupplemental issue I was
surprised at the diversity of AIapplications that are described
.
So you know, many veterinariansmight be familiar with things
like AI scribes, which arehelped to complete your clinical
(09:10):
note, and in fact Cornell juststarted using them recently as
well.
But we may not realize that AIis also used in diagnostic tests
or diagnostic algorithms.
So, for example, there's apaper describing AI-assisted
stethoscopes and motion sensorsto detect bovine respiratory
disease.
Veterinarians may also not knowhow machine learning algorithms
(09:32):
are being used for prediction,so for example, predicting the
risk of specific diseases incats or predicting the risk of
Lyme disease, and then how wecan use that information to help
target our preventive measuresor our diagnostic testing.
And then I think also what'ssurprising is that these
applications are not just usedlike on the clinic floor, but
(09:52):
also to really accelerateveterinary research.
So, for example, there's apaper from my lab about how we
can fill in missing data fromnational antimicrobial
resistance surveillance datasets using machine learning
methods.
And if we can accelerateveterinary research, we'll have
a lot more exciting things toshare and improve our veterinary
medicine.
So I think that you know,regardless of what your interest
(10:15):
is, all of your AJVR readerswill hopefully find an AI
application that they think isrelevant to them in that special
issue.
Lisa Fortier (10:24):
Yeah, fantastic.
And to the readers andlisteners it's open access, so
anybody can read it anywhere.
You don't have to subscribe.
Sarah and I, weekly on thispodcast, we ask the people that
we're interviewing, who arealways authors on manuscripts do
you see a role for AI ormachine learning in your area?
And all of them come up withsome fascinating responses.
(10:48):
We recently had someone talkingabout ice baths for horses in
the development phase oflaminitis and they were like,
yes, putting all this researchtogether and all of these
clinical parameters, so maybe wecan noodle on that together as
a team.
And how can we come up withthose, Because those are
clinicians and scientists andhow can we get them involved in
(11:10):
these multidisciplinary teams aswell?
So maybe we can, Sarah and Ican harvest all those responses
and share them at your nextsymposium for people to find
other clinical areas to harness,because that data is out there.
Casey Cazer (11:26):
That would be really exciting because you're
right, like we need differentcollaborations to come together
to bring together the data andthe skills and the expertise to
make these things happen.
Renata Ivanek (11:36):
And I would add to that we also need problems
that are worth solving, orproblems where there is strong
demand to be solved.
Lisa Fortier (11:44):
You're on it Sarah .
Sarah Wright (11:45):
I was going to say , I see a spreadsheet forming in
my mind right now, so staytuned after this for that, but I
was actually just reading toorecently they're even looking at
AI to help, like, diagnosecertain diseases in corals,
which is like obviously a reallybig deal, especially in the
Caribbean right now.
So, it's fascinating how muchit can do.
For those of you just joiningus, we're discussing the AJVR AI
(12:08):
supplemental issue with ourguests Casey Parminder and
Renata.
So, Renata, what are the nextsteps for research in AI
veterinary medicine?
Renata Ivanek (12:17):
Oh, I think we are into for a lot, a lot more
exciting discoveries, and I'lltalk just for about a few that
are within my own little circleof research.
So I think there will be a lotof new developments in
development of data systems thatare able to retrieve and store
and maybe share in aprivacy-protected manner, data,
(12:40):
because data and alsostandardization of that data,
because if data AI needs a lotof data and if data is not
standardized, if thispractitioner and this
practitioner are using differentwords, different criteria for
diagnosis, then it absolutelymakes no sense to make any
insights.
It doesn't matter how capablethat AI is, the results will be
(13:01):
completely meaningless.
Another area where I thinkthere will be a lot of movement
will be predictive analytics, incase you already mentioned that
a little bit.
So, because we are veryinterested in using AI to
predict disease progression ordisease outcomes or patient
outcomes, so that we canidentify individuals that are at
(13:23):
higher risk of developingdisease, for example diabetes,
to start to treat them earlieror maybe slow down progression
of disease so that the diseasedoesn't actually even develop
fully.
For example, in my lab, we arevery interested in infectious
disease dynamics and how we canpredict infection spread.
So, in all times we wouldhandcraft equations to describe
(13:48):
the dynamics and then we willstudy them.
That takes a lot of effort,expertise and synthesis a lot of
information.
Now we are using, or we want touse, AI to learn infection
dynamics directly from the data.
Another area that is ofinterest how do you predict
where is threat?
And then again, can you usesome pattern of movement or
(14:10):
pattern of sensors?
We are, for example, developinga new technology that will
predict where is contaminationin complex environments like
healthcare settings more likelyContamination, for example, with
nosocomial pathogens and sotrying to make that a little
more efficient.
Another area where I think therewill be a lot of movement and
we actually have already seenmovement is surveillance and
(14:33):
monitoring, for example, onfarms.
Nowadays it's becoming more andmore common that farmers and
veterinarians can monitoranimals, thanks to wearables and
different sensors, remotely,and they can get alerts when an
animal is in distress or needshelp.
And because AI technology isable to take data from various
(14:56):
resources, which can includeveterinary records, but a lot
more data, all kinds of sensorsor cameras or audio records to
find patterns and provide uswith early warning systems.
So, for example, in ourresearch we are developing an
AI-supported tool forantimicrobial stewardship in
(15:19):
livestock that would providefarmers with insights to improve
their business and animalhealth and welfare, while at the
same time securing that we cancontinue using antimicrobials
for the future.
So a lot more is coming.
It's a really really, reallydynamic field.
Right now, it's kind of quiteexciting.
Lisa Fortier (15:40):
Yeah, thanks for all those great examples.
Renata, if you had advice togive to a veterinarian or a
veterinary student who'sinterested in learning more
about the intersection of AI andveterinary medicine, what
advice would you give them?
Renata Ivanek (15:54):
Yeah, I think a big one is about data quality.
So we already alluded to this,and the old saying garbage in
garbage out definitely holds forAI technologies.
And so I would say, if you area practitioner or you have your
(16:18):
future practitioner, think, tryto understand what kind of data
are used to train or even tomake prediction with this AI.
How were data collected to makesure that AI insight is based
on reliable information?
Another one would be don't rushinto this, into integration.
If, when you are picking orwhen you are selecting air
(16:40):
technology that you want tointegrate into your practice or
your everyday work, pick thosefirst that there have a lot of
research and testing behind them, and then, when you start using
them again, don't rush.
Give yourself time to learn howto use them to full potential
and make mistakes.
Make mistakes in a safe spacebefore you integrate fully into
(17:04):
their use.
Also, so this AI.
Why we would use AI?
Because they would make ourlives better and also the
disease outcomes better.
So we definitely see reason touse them good rationale but at
what cost?
And for example and I'm nottalking about monetary costs,
(17:25):
I'm talking about what wouldthat do to relationship, for
example, your relationship withthe client, and then find maybe
one other piece of advice Don'tforce it.
If you are using, if you see anAI technology and it just
doesn't sit right with you, it'squite likely others, other
practitioners, will not like iteither.
(17:47):
And just wait, Market willprovide either the same
technology, new features or abetter technology that will fit
exactly what your needs are.
And then, finally this isreally as we already alluded
this is such a fast-paced field,and being informed will be
(18:08):
important and difficult at thesame time, and so keeping up
with new advancements, both inAI and veterinary medicine, will
be really important, and we cando that, All of us can do that.
Journals, conferences, we cantake online courses just to be
able to better understand how wecan use this technology.
(18:28):
These are all good resources,and even this podcast and this
special issue is a step in theright direction to keep informed
and open-minded.
Sarah Wright (18:40):
Yeah, very good advice, thank you.
Now, this next question isdefinitely one of our more
challenging questions.
If you can take all theinformation that's in this
supplemental issue and boil itdown to one really important
nugget of take-home informationthat's going to help
veterinarians, what should theyknow?
So, casey, what is one piece ofinformation the veterinarian
should know about the AJVR-AIsupplemental issue?
Casey Cazer (19:04):
That is really difficult and I'm going to kind
of tack on to what Renata wassaying about good data.
So we need veterinarians torealize that to have high
quality AI tools and useful AItools, you need a lot of high
quality data.
(19:36):
Everything in your practicemanagement software or your farm
management software, there's alot of information, but that
data is often either low qualityor isolated and difficult to
connect to other bits ofinformation.
And, as Renata said, there's asaying garbage in, garbage out.
So if you don't have good data,you're not going to get good
results.
So what I think veterinarianscan actually do quite a lot to
improve data quality andtherefore have better AI tools.
So, for example, many of thenew AI tools for small animal
veterinarians are being builtfrom clinical notes right, the
(19:57):
information that we put in aboutthe patients that we see, and
so we can make sure that yourclinical notes are accurate and
complete, including usefulthings like either master
problem lists or diagnosis codes, so that those AI tools can use
that to identify relevant cases.
And in fact, there's a littlebit of an AI loop here where you
(20:18):
can use an AI scribe to helpwrite your clinical note that's
listening to your appointmentand therefore write a better and
more complete clinical note.
Sarah Wright (20:27):
And the other side of the relationship.
What's one thing that clientsshould know about this important
supplemental issue?
Parminder Basran (20:33):
Well, I'm going to defer to you, Renata,
about what that students shouldprobably know about this.
But the one small thing that Iwould add in relation to things
to take away from thesupplemental issue is is this
the breadth and scope of thekinds of things that are
presented in the supplementitself demonstrates to me how
big of a tent veterinarymedicine is in general and how
(20:56):
AI fits inside this giant tentof different specialties.
And so, because of that,there's a unique advantage in
veterinary medicine in the sensethat it's easy to survey and
essay the different adoptions ofartificial intelligence in
different disciplines and reallythink about how one might be
able to transfer that kind ofinformation or application into
(21:20):
their domain, and that, I think,is a really exciting thing to
have.
Often, if you're a specialistwithin a specialty, you don't
get a chance to see a lot of thedifferent applications and
different disciplines.
So I think that's the one thingthat I personally would take
away from this is that it reallyis just an excellent example of
(21:42):
how big the tent is and howmuch we can learn from each
other.
Renata Ivanek (21:48):
Okay, so I would piggyback on that and I would
focus on ethics, and the reasonfor that is because ethics is
actually still AI, ethics isstill under development.
We really still don't haveguidelines or rules how we
should apply mindfully AItechnologies and we still
actually don't even know whatthe full breadth of even
benefits or dangers, everythingthat can go wrong.
(22:11):
And so, as we are tapping intothis new technology, I think we
all have to take responsibilityto think about AI applications,
about what's ethical, what'smoral.
So a lot is more coming, but Ithink there are still some
things that we know will holdlike in any future guidelines.
(22:32):
First, data privacy and securityhas to be preserved and it has
to be clear whether the clientprovided informed consent that
data could be used for AI.
We know from tragedy, fromhallucination we hear about,
that AI system can have inherentbiases, so we have to accept
the possibility that AI could bewrong, it doesn't matter how
(22:55):
sophisticated it could be wrong.
Also, we have to be able tomake an effort to understand how
did AI come to a conclusion ordecision, especially if that
decision contradicts our own,and this is important to explain
to ourselves what should wetrust?
But also to be able to explainthat to clients.
And then I think, maybe aboveall, veterinarians are top
(23:19):
experts in this field, and sothey should retain ultimate
responsibility for patient care,even when AI is used.
Lisa Fortier (23:27):
Yeah, really, really great points.
Thank you guys again for beinghere.
We learned a ton reading allthe manuscripts, even more today
, so thank you again for yourtime.
Thank you.
Sarah Wright (23:38):
Thank you.
Lisa Fortier (23:40):
As we wind down, we like to ask a little bit of a
fun question.
So, Casey, we'll start with you.
When you complete a puzzle, doyou begin with the interior,
middle or the exterior borderpieces begin?
Casey Cazer (23:52):
So, I actually I like to do puzzles, but I'll do
them like over Christmas andthat's it Cause otherwise I'm
like a puzzle addict and I'lljust like doing them.
And I used to always start withthe border and recently I've
been starting with the middle,like I'll sort them and find
some interesting feature in themiddle and start with that
Interesting.
Lisa Fortier (24:11):
We haven't had a switcher yet.
Our prediction is we're goingto retrospectively look at this,
but typically we find thatsurgeon types do the exterior
and more medicine-leaning peopledo the interior.
So maybe your communitypractice is getting to you.
Casey Cazer (24:31):
Maybe you could build an AIM-I model to see if
you can predict that feature.
Lisa Fortier (24:35):
Okay, back to you, Renata.
If you could have a superpower,what would it be, and why?
Renata Ivanek (24:41):
I want to be a healer.
I feel it's like a catch-all.
I could then treat and saveeverybody around me, whether
they're and me included, andmaybe what helps with living
long and happy life.
Lisa Fortier (24:55):
Good, Best politics start at home, so heal
yourself first.
Parminder what is the oldest orthe most interesting thing on
your desk or in your desk drawer?
Parminder Basran (25:05):
Well, I was going to say something you know
cheesy, like my savvy mug, um uh, just to just to promote that.
But I have what I write, what Ihave on my hand.
I'm sorry it's not really allin focus, but this is a piece of
carbon.
Do you remember the showoppenheimer?
yes you remember there was ascene in a basement in the
university of chicago where theywere testing for the very first
(25:28):
time a controlled nuclearfusion or fission.
I don't remember the scene, butI remember the book.
There was a scene down there.
So the very first nuclearreaction.
What took place underneath theUniversity of Chicago?
It was all contained with largeslabs of carbon to help shield
the radiation.
(25:49):
This is an actual slab from thePio-1 nuclear reactor.
So this is a very nerdy thing.
I love this thing and I bragabout it to my PhD supervisor
every time I see him.
Lisa Fortier (26:04):
That is super cool .
I'm a trickster, so if I gotinto your office I would make a
fake piece and see how long ittook you to notice.
Parminder Basran (26:13):
I've got lots of really weird things in my
office.
I've got Lego pieces and it's a.
It's a I love.
I love collecting odd weirdthings.
Casey Cazer (26:25):
So 2.0, the Symposium on Artificial
Intelligence and VeterinaryMedicine.
It's going to happen at CornellUniversity, may 16th to the
18th 2025.
You can get more information atcornelaivetorg.
We're going to have somefantastic keynotes about
wildlife medicine, livestockmedicine and companion animal
medicine, and a little bit onhuman health and how AI is
(26:48):
influencing all of those areas.
Sarah Wright (26:51):
And to our listeners.
We'll have the link to thatwebsite as part of the
description of this episode, soif you scroll down, you can find
it.
Well, thank you so much again.
Parminder, casey and RenataReally appreciate you being here
today serving as guest editorsfor this supplemental issue and
just sharing insights about itto our listeners.
Thanks very much.
Parminder Basran (27:11):
Thank you.
Sarah Wright (27:12):
And again to our listeners.
You can read the AIsupplemental issue in AJVR.
I'm Sarah Wright with LisaFortier.
Be on the lookout for nextweek's episode and don't forget
to leave us a rating and reviewon Apple Podcasts or whatever
platform you listen to.
Mealtime is more than a bowl of food.
It's the foundation for a longand healthy life.
At Wellness, dry recipes arecrafted with the highest quality
natural ingredients for a tastethat pets love and
scientifically proven throughAFCO feeding trials to ensure
that every complete and balancedmeal supports the five signs of
wellbeing.
With Wellness Pet Foods, petparents can share a life of
(00:25):
well-being together with theirpet.
Sarah Wright (00:31):
This is Veterinary Vertex, a podcast of the AVMA
Journals.
In this episode we chat aboutthe March AJVR AI Supplemental
Issue From Bark to BytesArtificial Intelligence
Transforming Veterinary Medicine, with our guests Casey Cazer,
Parminder Basran and RenataIvanek.
Lisa Fortier (00:50):
Welcome everyone.
I'm Editor-in-Chief LisaFortier, with co-host and
Associate Editor Sarah Wright.
Parminder and Casey welcomeback.
Renata thank you so much forbeing here with us today.
Great to see my Cornellcolleagues.
Renata Ivanek (01:03):
Happy to be here, thank you.
Parminder Basran (01:04):
Yes, absolutely.
Sarah Wright (01:06):
All right, let's dive right in and talk all about
AI.
So, you are the guest editorsfor the AJVR supplemental issue.
From Bark to Bytes ArtificialIntelligence, transforming
Veterinary Medicine.
Casey, what can readers expectto take away from these articles
?
Casey Cazer (01:26):
In this special issue, I think readers will find
papers written by differentleading experts and researchers
who are using AI in veterinarymedicine.
These authors were actuallypresenters and attendees at the
first annual Symposium ofArtificial Intelligence in
Veterinary Medicine that washeld at Cornell University in
April 2024.
So, whether you are unsure whatan LLM is or whether you're
using AI tools every day in yourveterinary practice, I think
(01:49):
there's something new foreveryone to learn in this
special issue.
I would encourage readers tostart with the first two
articles in the issue, whichprovide sort of a gentle
introduction to AI in vet medand also the FDA's regulatory
perspective on AI and machinelearning in veterinary medicine,
and the takeaways that I tookfrom these two articles is that
(02:10):
AI can be very useful right nowor in the future in many aspects
of veterinary medicine,including diagnostics, drug
development, clinical decisionmaking, record keeping and
client communication.
However, there are a lot ofpotential pitfalls, including a
need for education ofveterinarians and clients on
these AI tools, a lack of highquality data to train the
(02:32):
machine learning models, andethical and trustworthiness
challenges.
Lisa Fortier (02:38):
Yeah, it was really fascinating.
I enjoyed reading every one ofthe articles.
I'm not honest, I didn'tunderstand all of them, but I
certainly learned so muchreading them.
Just the breadth of how AI isalready involved in so much and
what the potentials are.
It's really fascinating.
So thank you for the idea toform this supplemental issue.
Casey Cazer (03:00):
Sure, so I hope that if people are interested in
this topic, they want toconsider attending the Symposium
for Artificial Intelligence inVeterinary Medicine 2.0, which
is going to be held at Cornellon May 16th to 18th 2025.
And you can get moreinformation on it at
cornellaivet.
org.
Lisa Fortier (03:20):
Perfect, that's when our forsythia is in full
bloom.
So come on out to Ithaca.
Hey, Parminder, what sparkedyour interest in artificial
intelligence and then with yourcolleagues Casey and Renata,
forming this symposium?
Parminder Basran (03:36):
When I first joined Cornell about almost six
years ago, there had alreadybeen a fair bit of research in
human medicine and artificialintelligence that I was
undertaking working in a humanhospital and when I arrived here
I realized that there wasn'tnearly the same amount of push
(03:57):
or, I guess, projects and theamount of resources that people
are spending in veterinarymedicine and artificial
intelligence.
So that to me reallydemonstrated sort of an unmet
need needed within ourenvironment here in veterinary
(04:18):
medicine.
So, there was a natural sort ofprogression of exploring
artificial intelligencemigrating from human medicine to
veterinary medicine.
And then soon after that, it'samazing how long ago two years
feels from now.
But if we were to go back in atime machine and think about
what life was like two years ago, there were tremendous
(04:46):
questions that we were askingourselves about artificial
intelligence and veterinarymedicine.
You know, things like ChatGPTweren't even on the radar and so
at that time a bunch of us gottogether and said, hey it would
be great if we could create anenvironment where we can bring
together people from differentdisciplines in human medicine
and in epidemiology and incompanion animal medicine and
(05:11):
computing science and engineersand just really talk about what
artificial intelligence means inthis space, and so that was the
germ of the idea is for us toreally just start having a
conversation about this a coupleof two years ago, and what
happened after that was we wereable to secure some funding
(05:34):
through the NIH and the FDA tosponsor the event, brought
together sort of these fourpillars of veterinary medicine
companion animal, populationmedicine, comparative health
basically bringing togetherdifferent components of
artificial intelligence andveterinary medicine in a single
(05:56):
setting.
We also realized at the timethat there weren't a lot of
avenues for our trainees to gettogether and talk about
artificial intelligence inveterinary medicine.
There may be a lot ofspecialized AI conferences, but
there wasn't really anythingthat really captured veterinary
medicine.
So again, trying to see if wecould create a safe space for
(06:21):
people of different disciplinesto get together and talk about
these important things wasreally important.
And so having that funding fromthe FDA and the NIH helped
really push the symposium alongand we were really happy about
that.
And we were very pleased thatwe helped kind of break down
some of the barriers forknowledge and education in that
(06:41):
process as well.
So you know, the more that wecan share knowledge about things
that we don't really understand, the less you know, the less
fearful we are of thattechnology, and so we were
pretty happy that we were ableto do this thing and, based on
our success of that event, it'sreally evolved and become
(07:06):
something of a beast of itself,so they're pretty happy about
that.
Lisa Fortier (07:17):
Yep.
Success comes with someresponsibility, then, doesn't it
?
Yeah, walking around at theposters and reading about the
speakers, how diverse theirbackgrounds were in coming and
their disciplines, and how it'sthe most collaborative field of
all it was fascinating.
(07:38):
Well done to get that sort ofattention as well on your first
one!
Parminder Basran (07:40):
I think that's a great point in the sense that
the thing that we know aboutartificial intelligence and
having successful practicalapplications of artificial
intelligence is that it can'tjust be done in a silo.
A single person can't do, can'tget success, uh by by, you know
, going alone.
So so bringing people withdifferent expertise and
(08:03):
disciplines together was a bigpart of what we wanted to do,
and indeed that really holds upin terms of the literature, in
terms of the successfulness ofAI adoption and implementation
in clinical practices, is thatthose environments where that
have lots of multidisciplinaryteams, people from different
backgrounds, tend to do verywell in deploying AI in their
(08:27):
communities and environments.
Lisa Fortier (08:28):
That's fabulous, Casey.
Back to you for a minute.
Sarah asked you a littleearlier what were some of the
interesting take-home messagesfor folks, but always, at least
for me, there was lots ofsurprises, both at the symposium
and in things I just didn'tknow.
What things surprised you inthe supplemental issue, what did
you learn that was like huh, Ididn't know that?
Casey Cazer (08:49):
Well, I think, actually similar to you and, how
you said, walking around thesymposium and looking at the
posters, you were surprised athow much diversity there was.
I think that in thesupplemental issue I was
surprised at the diversity of AIapplications that are described
.
So you know, many veterinariansmight be familiar with things
like AI scribes, which arehelped to complete your clinical
(09:10):
note, and in fact Cornell juststarted using them recently as
well.
But we may not realize that AIis also used in diagnostic tests
or diagnostic algorithms.
So, for example, there's apaper describing AI-assisted
stethoscopes and motion sensorsto detect bovine respiratory
disease.
Veterinarians may also not knowhow machine learning algorithms
(09:32):
are being used for prediction,so for example, predicting the
risk of specific diseases incats or predicting the risk of
Lyme disease, and then how wecan use that information to help
target our preventive measuresor our diagnostic testing.
And then I think also what'ssurprising is that these
applications are not just usedlike on the clinic floor, but
(09:52):
also to really accelerateveterinary research.
So, for example, there's apaper from my lab about how we
can fill in missing data fromnational antimicrobial
resistance surveillance datasets using machine learning
methods.
And if we can accelerateveterinary research, we'll have
a lot more exciting things toshare and improve our veterinary
medicine.
So I think that you know,regardless of what your interest
(10:15):
is, all of your AJVR readerswill hopefully find an AI
application that they think isrelevant to them in that special
issue.
Lisa Fortier (10:24):
Yeah, fantastic.
And to the readers andlisteners it's open access, so
anybody can read it anywhere.
You don't have to subscribe.
Sarah and I, weekly on thispodcast, we ask the people that
we're interviewing, who arealways authors on manuscripts do
you see a role for AI ormachine learning in your area?
And all of them come up withsome fascinating responses.
(10:48):
We recently had someone talkingabout ice baths for horses in
the development phase oflaminitis and they were like,
yes, putting all this researchtogether and all of these
clinical parameters, so maybe wecan noodle on that together as
a team.
And how can we come up withthose, Because those are
clinicians and scientists andhow can we get them involved in
(11:10):
these multidisciplinary teams aswell?
So maybe we can, Sarah and Ican harvest all those responses
and share them at your nextsymposium for people to find
other clinical areas to harness,because that data is out there.
Casey Cazer (11:26):
That would be really exciting because you're
right, like we need differentcollaborations to come together
to bring together the data andthe skills and the expertise to
make these things happen.
Renata Ivanek (11:36):
And I would add to that we also need problems
that are worth solving, orproblems where there is strong
demand to be solved.
Lisa Fortier (11:44):
You're on it Sarah .
Sarah Wright (11:45):
I was going to say , I see a spreadsheet forming in
my mind right now, so staytuned after this for that, but I
was actually just reading toorecently they're even looking at
AI to help, like, diagnosecertain diseases in corals,
which is like obviously a reallybig deal, especially in the
Caribbean right now.
So, it's fascinating how muchit can do.
For those of you just joiningus, we're discussing the AJVR AI
(12:08):
supplemental issue with ourguests Casey Parminder and
Renata.
So, Renata, what are the nextsteps for research in AI
veterinary medicine?
Renata Ivanek (12:17):
Oh, I think we are into for a lot, a lot more
exciting discoveries, and I'lltalk just for about a few that
are within my own little circleof research.
So I think there will be a lotof new developments in
development of data systems thatare able to retrieve and store
and maybe share in aprivacy-protected manner, data,
(12:40):
because data and alsostandardization of that data,
because if data AI needs a lotof data and if data is not
standardized, if thispractitioner and this
practitioner are using differentwords, different criteria for
diagnosis, then it absolutelymakes no sense to make any
insights.
It doesn't matter how capablethat AI is, the results will be
(13:01):
completely meaningless.
Another area where I thinkthere will be a lot of movement
will be predictive analytics, incase you already mentioned that
a little bit.
So, because we are veryinterested in using AI to
predict disease progression ordisease outcomes or patient
outcomes, so that we canidentify individuals that are at
(13:23):
higher risk of developingdisease, for example diabetes,
to start to treat them earlieror maybe slow down progression
of disease so that the diseasedoesn't actually even develop
fully.
For example, in my lab, we arevery interested in infectious
disease dynamics and how we canpredict infection spread.
So, in all times we wouldhandcraft equations to describe
(13:48):
the dynamics and then we willstudy them.
That takes a lot of effort,expertise and synthesis a lot of
information.
Now we are using, or we want touse, AI to learn infection
dynamics directly from the data.
Another area that is ofinterest how do you predict
where is threat?
And then again, can you usesome pattern of movement or
(14:10):
pattern of sensors?
We are, for example, developinga new technology that will
predict where is contaminationin complex environments like
healthcare settings more likelyContamination, for example, with
nosocomial pathogens and sotrying to make that a little
more efficient.
Another area where I think therewill be a lot of movement and
we actually have already seenmovement is surveillance and
(14:33):
monitoring, for example, onfarms.
Nowadays it's becoming more andmore common that farmers and
veterinarians can monitoranimals, thanks to wearables and
different sensors, remotely,and they can get alerts when an
animal is in distress or needshelp.
And because AI technology isable to take data from various
(14:56):
resources, which can includeveterinary records, but a lot
more data, all kinds of sensorsor cameras or audio records to
find patterns and provide uswith early warning systems.
So, for example, in ourresearch we are developing an
AI-supported tool forantimicrobial stewardship in
(15:19):
livestock that would providefarmers with insights to improve
their business and animalhealth and welfare, while at the
same time securing that we cancontinue using antimicrobials
for the future.
So a lot more is coming.
It's a really really, reallydynamic field.
Right now, it's kind of quiteexciting.
Lisa Fortier (15:40):
Yeah, thanks for all those great examples.
Renata, if you had advice togive to a veterinarian or a
veterinary student who'sinterested in learning more
about the intersection of AI andveterinary medicine, what
advice would you give them?
Renata Ivanek (15:54):
Yeah, I think a big one is about data quality.
So we already alluded to this,and the old saying garbage in
garbage out definitely holds forAI technologies.
And so I would say, if you area practitioner or you have your
(16:18):
future practitioner, think, tryto understand what kind of data
are used to train or even tomake prediction with this AI.
How were data collected to makesure that AI insight is based
on reliable information?
Another one would be don't rushinto this, into integration.
If, when you are picking orwhen you are selecting air
(16:40):
technology that you want tointegrate into your practice or
your everyday work, pick thosefirst that there have a lot of
research and testing behind them, and then, when you start using
them again, don't rush.
Give yourself time to learn howto use them to full potential
and make mistakes.
Make mistakes in a safe spacebefore you integrate fully into
(17:04):
their use.
Also, so this AI.
Why we would use AI?
Because they would make ourlives better and also the
disease outcomes better.
So we definitely see reason touse them good rationale but at
what cost?
And for example and I'm nottalking about monetary costs,
(17:25):
I'm talking about what wouldthat do to relationship, for
example, your relationship withthe client, and then find maybe
one other piece of advice Don'tforce it.
If you are using, if you see anAI technology and it just
doesn't sit right with you, it'squite likely others, other
practitioners, will not like iteither.
(17:47):
And just wait, Market willprovide either the same
technology, new features or abetter technology that will fit
exactly what your needs are.
And then, finally this isreally as we already alluded
this is such a fast-paced field,and being informed will be
(18:08):
important and difficult at thesame time, and so keeping up
with new advancements, both inAI and veterinary medicine, will
be really important, and we cando that, All of us can do that.
Journals, conferences, we cantake online courses just to be
able to better understand how wecan use this technology.
(18:28):
These are all good resources,and even this podcast and this
special issue is a step in theright direction to keep informed
and open-minded.
Sarah Wright (18:40):
Yeah, very good advice, thank you.
Now, this next question isdefinitely one of our more
challenging questions.
If you can take all theinformation that's in this
supplemental issue and boil itdown to one really important
nugget of take-home informationthat's going to help
veterinarians, what should theyknow?
So, casey, what is one piece ofinformation the veterinarian
should know about the AJVR-AIsupplemental issue?
Casey Cazer (19:04):
That is really difficult and I'm going to kind
of tack on to what Renata wassaying about good data.
So we need veterinarians torealize that to have high
quality AI tools and useful AItools, you need a lot of high
quality data.
(19:36):
Everything in your practicemanagement software or your farm
management software, there's alot of information, but that
data is often either low qualityor isolated and difficult to
connect to other bits ofinformation.
And, as Renata said, there's asaying garbage in, garbage out.
So if you don't have good data,you're not going to get good
results.
So what I think veterinarianscan actually do quite a lot to
improve data quality andtherefore have better AI tools.
So, for example, many of thenew AI tools for small animal
veterinarians are being builtfrom clinical notes right, the
(19:57):
information that we put in aboutthe patients that we see, and
so we can make sure that yourclinical notes are accurate and
complete, including usefulthings like either master
problem lists or diagnosis codes, so that those AI tools can use
that to identify relevant cases.
And in fact, there's a littlebit of an AI loop here where you
(20:18):
can use an AI scribe to helpwrite your clinical note that's
listening to your appointmentand therefore write a better and
more complete clinical note.
Sarah Wright (20:27):
And the other side of the relationship.
What's one thing that clientsshould know about this important
supplemental issue?
Parminder Basran (20:33):
Well, I'm going to defer to you, Renata,
about what that students shouldprobably know about this.
But the one small thing that Iwould add in relation to things
to take away from thesupplemental issue is is this
the breadth and scope of thekinds of things that are
presented in the supplementitself demonstrates to me how
big of a tent veterinarymedicine is in general and how
(20:56):
AI fits inside this giant tentof different specialties.
And so, because of that,there's a unique advantage in
veterinary medicine in the sensethat it's easy to survey and
essay the different adoptions ofartificial intelligence in
different disciplines and reallythink about how one might be
able to transfer that kind ofinformation or application into
(21:20):
their domain, and that, I think,is a really exciting thing to
have.
Often, if you're a specialistwithin a specialty, you don't
get a chance to see a lot of thedifferent applications and
different disciplines.
So I think that's the one thingthat I personally would take
away from this is that it reallyis just an excellent example of
(21:42):
how big the tent is and howmuch we can learn from each
other.
Renata Ivanek (21:48):
Okay, so I would piggyback on that and I would
focus on ethics, and the reasonfor that is because ethics is
actually still AI, ethics isstill under development.
We really still don't haveguidelines or rules how we
should apply mindfully AItechnologies and we still
actually don't even know whatthe full breadth of even
benefits or dangers, everythingthat can go wrong.
(22:11):
And so, as we are tapping intothis new technology, I think we
all have to take responsibilityto think about AI applications,
about what's ethical, what'smoral.
So a lot is more coming, but Ithink there are still some
things that we know will holdlike in any future guidelines.
(22:32):
First, data privacy and securityhas to be preserved and it has
to be clear whether the clientprovided informed consent that
data could be used for AI.
We know from tragedy, fromhallucination we hear about,
that AI system can have inherentbiases, so we have to accept
the possibility that AI could bewrong, it doesn't matter how
(22:55):
sophisticated it could be wrong.
Also, we have to be able tomake an effort to understand how
did AI come to a conclusion ordecision, especially if that
decision contradicts our own,and this is important to explain
to ourselves what should wetrust?
But also to be able to explainthat to clients.
And then I think, maybe aboveall, veterinarians are top
(23:19):
experts in this field, and sothey should retain ultimate
responsibility for patient care,even when AI is used.
Lisa Fortier (23:27):
Yeah, really, really great points.
Thank you guys again for beinghere.
We learned a ton reading allthe manuscripts, even more today
, so thank you again for yourtime.
Thank you.
Sarah Wright (23:38):
Thank you.
Lisa Fortier (23:40):
As we wind down, we like to ask a little bit of a
fun question.
So, Casey, we'll start with you.
When you complete a puzzle, doyou begin with the interior,
middle or the exterior borderpieces begin?
Casey Cazer (23:52):
So, I actually I like to do puzzles, but I'll do
them like over Christmas andthat's it Cause otherwise I'm
like a puzzle addict and I'lljust like doing them.
And I used to always start withthe border and recently I've
been starting with the middle,like I'll sort them and find
some interesting feature in themiddle and start with that
Interesting.
Lisa Fortier (24:11):
We haven't had a switcher yet.
Our prediction is we're goingto retrospectively look at this,
but typically we find thatsurgeon types do the exterior
and more medicine-leaning peopledo the interior.
So maybe your communitypractice is getting to you.
Casey Cazer (24:31):
Maybe you could build an AIM-I model to see if
you can predict that feature.
Lisa Fortier (24:35):
Okay, back to you, Renata.
If you could have a superpower,what would it be, and why?
Renata Ivanek (24:41):
I want to be a healer.
I feel it's like a catch-all.
I could then treat and saveeverybody around me, whether
they're and me included, andmaybe what helps with living
long and happy life.
Lisa Fortier (24:55):
Good, Best politics start at home, so heal
yourself first.
Parminder what is the oldest orthe most interesting thing on
your desk or in your desk drawer?
Parminder Basran (25:05):
Well, I was going to say something you know
cheesy, like my savvy mug, um uh, just to just to promote that.
But I have what I write, what Ihave on my hand.
I'm sorry it's not really allin focus, but this is a piece of
carbon.
Do you remember the showoppenheimer?
yes you remember there was ascene in a basement in the
university of chicago where theywere testing for the very first
(25:28):
time a controlled nuclearfusion or fission.
I don't remember the scene, butI remember the book.
There was a scene down there.
So the very first nuclearreaction.
What took place underneath theUniversity of Chicago?
It was all contained with largeslabs of carbon to help shield
the radiation.
(25:49):
This is an actual slab from thePio-1 nuclear reactor.
So this is a very nerdy thing.
I love this thing and I bragabout it to my PhD supervisor
every time I see him.
Lisa Fortier (26:04):
That is super cool .
I'm a trickster, so if I gotinto your office I would make a
fake piece and see how long ittook you to notice.
Parminder Basran (26:13):
I've got lots of really weird things in my
office.
I've got Lego pieces and it's a.
It's a I love.
I love collecting odd weirdthings.
Casey Cazer (26:25):
So 2.0, the Symposium on Artificial
Intelligence and VeterinaryMedicine.
It's going to happen at CornellUniversity, may 16th to the
18th 2025.
You can get more information atcornelaivetorg.
We're going to have somefantastic keynotes about
wildlife medicine, livestockmedicine and companion animal
medicine, and a little bit onhuman health and how AI is
(26:48):
influencing all of those areas.
Sarah Wright (26:51):
And to our listeners.
We'll have the link to thatwebsite as part of the
description of this episode, soif you scroll down, you can find
it.
Well, thank you so much again.
Parminder, casey and RenataReally appreciate you being here
today serving as guest editorsfor this supplemental issue and
just sharing insights about itto our listeners.
Thanks very much.
Parminder Basran (27:11):
Thank you.
Sarah Wright (27:12):
And again to our listeners.
You can read the AIsupplemental issue in AJVR.
I'm Sarah Wright with LisaFortier.
Be on the lookout for nextweek's episode and don't forget
to leave us a rating and reviewon Apple Podcasts or whatever
platform you listen to.
Popular Podcasts
Boysober
Have you ever wondered what life might be like if you stopped worrying about being wanted, and focused on understanding what you actually want? That was the question Hope Woodard asked herself after a string of situationships inspired her to take a break from sex and dating. She went "boysober," a personal concept that sparked a global movement among women looking to prioritize themselves over men. Now, Hope is looking to expand the ways we explore our relationship to relationships. Taking a bold, unfiltered look into modern love, romance, and self-discovery, Boysober will dive into messy stories about dating, sex, love, friendship, and breaking generational patterns—all with humor, vulnerability, and a fresh perspective.
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com