The third episode of DigiPath Digest just took place live, but I have an audio version for the listeners.
DigiPath Digest is a review of digital pathology and IA publications abstract review that I host weekly as a live stream (on YouTube, LinkedIn, Facebook etc.)
Here is the video version if you learn more visually
Today the abstracts we discussed centered around innovations in disease detection and prognosis powered by digital pathology and AI.
TIMESTAMPS:
00:00 Welcome and Introduction
00:35 DigiPath Digest Overview
01:14 Engaging with the Audience
06:09 Abstract Review: AI in Liver Fibrosis
11:21 Abstract Review: AI in Prostate Cancer
16:43 Abstract Review: AI in Glioblastoma
23:02 Abstract Review: AI in Red Blood Cell Analysis
28:38 Upcoming Events and Announcements
34:18 Closing Remarks and Future Episodes
TODAY'S ABSTRACTS & RESOURCES:
Become a Digital Pathology Trailblazer get the "Digital Pathology 101" FREE E-book and join us!
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
Welcome my digital pathologytrailblazers.
Today was the third digit pathdigest abstract review.
And my friends, this is becominga thing.
I see people who are joiningover and over again.
And.
It's at 6:00 AM in the us.
So I know that people, well, no,I take the back because I had
(00:23):
one person who, uh, is joining.
I think he's he joined.
I already like the third door,the second time.
Anyway, so there are people inthe us.
Cool.
The morning.
So thank you so much.
And when I was reviewing theabstracts today, there is a
theme that these publicationsare not really from pathology
journals.
So I'm not going to be givingyou more intro because there's
(00:43):
intro already in the digitaldigest.
So let's dive into it.
Learn about the newest digitalpathology trends in science and
industry.
Meet the most interesting peoplein the niche, and gain insights
relevant to your own projects.
Here is where pathology meetscomputer science.
(01:04):
You are listening to the DigitalPathology Podcast with your
host, Dr.
Aleksandr Zhurav.
Aleksandra (01:14):
Good morning.
Good morning, everyone.
It's 6 a.
m.
in Pennsylvania, and I welcomeyou to the third edition of
DigiPath Digest.
When you log in here, let meknow when you are tuning in
(01:34):
from.
Give me a hi and when Whereyou're where you're at right
now.
So, um, let me tell you a story.
Of course.
I have my beautiful, fantasticnew camera here.
That didn't work.
So I'm just doing my normal, um,computer camera.
But let me know where you'retuning in from.
(01:57):
I see you're joining and let mejust say hello in the comments
and welcome you.
Um, we have a few abstractstoday.
Three.
Oh, Erica.
So great to have you here.
Hello.
(02:18):
And Robert, you made it fromMaine.
Yeah.
So there was a little bit of aglitch yesterday.
I think I had it initially foryesterday, um, at 6.
00 PM.
And, uh, it's always at 6 AM inthe morning, but, um, Robert
sent me a message that, Hey, isit yesterday or today it's
(02:39):
today.
And we have Jason fromBaltimore.
Baltimore is not that far fromme.
It's one and a half hours.
So fantastic to have you hereguys.
Um, I know there's going to bemore people joining, but for
those who already were here ontime, um, here agenda guys, I'm
serious.
Um, so.
For those who are here, butdon't know me, I'm Dr.
(03:00):
Alex Rudaf and I'm a veterinarypathologist and the founder of
Digital Pathology Place.
And I'm on a mission to teachyou all you need to know about
digital pathology.
And part of this mission is mybook.
If you didn't grab the book yet,there is one when you go to my
website, digitalpathologyplace.
com, you can download this onefor free and Let's dive into one
(03:25):
question before, um, because,uh, so now what I'm doing, I'm
reviewing the abstracts thatcome from PubMed alerts, but I
also, uh, discovered this, um,software for literature review,
literature research calledUndermined and I showed it to
(03:45):
you last time, but I'm going toshare it with you one more time
because I want to ask yousomething.
Step screen, unsure screen, uh,so undermined, you should see it
now.
So this is this, uh, researchengine.
(04:06):
So, um, what, um, last, thisweek I had a, uh, ergonomic
training and.
Welcome.
Ergonomics training at work.
And, uh, I was like, okay, I'mworking remotely.
So I do have some stuff.
I have like a standing desk.
I have a walking treadmill forthings because when I started
(04:30):
working from home and I have thewatch and I have the ring to
measure my activity, theactivity goes, uh, a lot.
Like it's so much less than whenI was actually driving to work
and I had to go from my car tosomewhere anyway.
And, uh, welcome Milena from theNetherlands.
Great to have you here anyway.
So, uh, long story short withthis tool undermined, you can do
(04:52):
research on specific topic.
And I did a research on, um,pathology, digital pathology
ergonomics.
So if this is something you'reinterested in me covering, I
mean, I'm going to cover this,um, At some point anyway, but if
you're interested in any othertopics, leave me a comment, what
topics you're interested in,like specific topics, like, I
(05:14):
don't know the scanners or, um,the AI models that are
available, uh, open source orsomething, whatever, uh, you're
interested in and hi to India.
And welcome, Ahmed from SaudiArabia.
Great to have you here, guys.
(05:34):
So any topic that you think, oh,I would need, like, I would want
something specific to that.
Not just what comes from PubMedevery week, but just specific to
that particular topic.
Let me know in the comments, thetopic.
And without further ado, let'smove to our abstracts.
(05:58):
Need to share a differentscreen, which is my tablet where
I can also draw, right?
Yes.
Okay.
Perfect.
So today the first topic is AIbased digital pathology provides
(06:19):
newer insights into lifestyleintervention induced fibrosis
regression in MASLD.
Thank you.
So, um, this is liver fibrosis.
I'm going to explain everythingto you, but, um, something
that's interesting, I'm lookingwhere these publications come
from, not only geographically.
So, um, this one is from China,but what kind of journals, uh,
(06:44):
they are in.
So this is like a liver internalmedicine journal.
Um, there are a few that wedon't have, uh, uh, In pathology
journals today, and they arefrom other disciplines where
people use digital pathologyand, and published there, which
I think is cool because justlike spreads digital pathology
from this niche of pathologists,because pathology is the gateway
(07:06):
to diagnosis.
And, um, that's why I'm happyabout that.
So.
What did they say here today?
Life stay intervention is themain stay of therapy for
metabolic dysfunctionassociated, uh, stato hepatitis.
So MASH, this is what we'retalking about.
(07:28):
And fibrosis is a keyconsequence that predicts
adverse clinical outcomes.
So when it goes very bad, youhave liver fibrosis.
I want to make it even bigger sothat you can see this.
Um, anyway, so what they didhere, they did imaging of
biopsies of pre treatment andpost treatment biopsies.
(07:48):
And I'm going to tell you whatthe treatment or like, um, what
the intervention was, um, butthey use a different method than
just our standard H and E.
They used harmonic generation tophoton excitation fluorescence
microscopy with, um, AI imageanalysis.
So, uh, it's, uh.
(08:08):
non staining method.
You excite the surface of thesample and you image it, you
image the fluorescence and themethod, um, to calculate the
level of fibrosis.
They called it Q fibrosis.
And this is, this providesqualitative assessment.
And so I thought that was alsointeresting because the non
(08:30):
staining methods are gettingmore popular.
prevalent.
And if you're just, if you'vejust joined because I see the
number of people joining andunjoining fluctuate, uh, be sure
to say hi and be sure to let meknow if there is a specific
topic you would like to cover inone of these, uh, journal clubs.
So, um, they examined thoseunstained sections.
(08:55):
So here's the keyword unstainedsections from paired liver
biopsies, baseline and end oftreatment intervention.
So pre and post, and they hadsomething that is called routine
lifestyle intervention andstrengthened lifestyle
intervention.
Um, And they actually sawfibrosis regression, which I
(09:16):
think is a big deal becausefibrosis is like, um, one of
those processes in pathologythat doesn't go away that
easily, but they saw it withthis lifestyle intervention.
Uh, and.
They basically say that usingdigital pathology, they could
detect more pronounced fibrosisregression.
(09:37):
Um, congratulations to them.
Let's see.
Where is it published?
And we already checked in theliver internal medicine.
Okay.
So that is our first abstractwithout staining.
I think this is cool because,um, so I told you the other day,
(09:59):
or I keep telling people, uh,that I was super excited about
the molecular predictions.
And for a long time, like 34years, uh, it was like only
academia, only research, nothinginto the clinic until there was
a publication in 2023, thatsomething actually made it in
(10:20):
the clinic.
The other thing that I'm superexcited about is the stainless
steel.
Uh, imaging direct to digitalimaging.
Sometimes you can do direct todigital from, um, from tissue,
like immediately without evenfixating the tissue or just like
cut it or freeze, cut it ordifferent methods.
(10:42):
And sometimes you need to fixit, but you don't have to stain
it.
So anyway, in each of these.
Uh, options, you skip some ofthe analog stacks, um, analog
steps of pathology of digitalpathology.
You go direct to digital and,and that's something that is
super, um, like a huge advantageof digital pathology because
(11:04):
radiology and all the otherdigital imaging, they have it.
They don't have to go to analogand then to digital like
pathology.
And I'm like, okay, until thatgoes away, uh, we're not going
to be as good as the other ones.
So Uh, we are as good as theother ones, uh, but just have
class slides.
Okay.
Now moving on to our nextabstract, uh, which is
(11:29):
artificial intelligence for, oh,sorry.
Artificial intelligence fordetection of prostate cancer in
biopsies during activesurveillance.
So, um, I assumed that Malepeople know what that is, but I
checked like, what's thedifference between active
surveys and surveillance andtreatment and basically active
(11:54):
surveillance is when you are nottreating, you're just monitoring
and you're checking.
You still do.
Um, biopsies pre and post atsome point, um, sorry, no, you
do the 1st biopsy and then youhave the results that are
telling you that there is nocancer or inconclusive for like,
low.
There is a specific Gleasonscore for that.
(12:15):
And, um.
That's active surveillance.
You don't do prostatectomy, youdon't treat, you just monitor.
And what happened here, they usean AI algorithm, which to me, it
was like, okay, um, well, weeven have an AI algorithm that
is, uh, cleared by the FDA forprostate.
(12:36):
So that's kind of like not thatnovel.
Let me check the journal again.
Because this one was also, Oh,so I didn't know this journal
BJU.
This is a British journal ofurology.
So this is again, a nonpathology journal that, um, is
(12:57):
actually publishing aboutdigital pathology and has the
keywords digital pathology.
So Um, they evaluated cancerdetect, uh, sorry to, uh, uh,
the objectives were to evaluatea cancer detecting artificial
intelligence algorithm on activesurveillance.
So they had, uh, and this is agroup from Norway, they had 180
(13:18):
patients in the, um, prostatecancer research, international
active surveillance priors.
Did I tell you that thepathology likes those, uh,
abbreviations?
Um, I think last time I toldyou.
So, uh, they had diagnostic.
So that the initial diagnosis,initial biopsy and let me zoom
(13:41):
in and their re biopsy slide.
So they had 2 biopsies and thedates were from 2011 to 2000.
20.
And they had a lot of thesebiopsies because they had 4,
000, um, 4, 744.
And they were scanned andanalyzed by an in house AI
(14:04):
cancer, uh, detection team.
Algorithm.
So they, they developed theirown, they didn't use any
commercial tool.
They just developed their own.
And what happened, uh, also theywere, uh, then, um, the goal
was, uh, that To analyze this,um, this algorithm for a
(14:25):
sensitivity specificity andaccuracy to predict the need for
active treatment.
So their surveillance and ifthere is a need for active
treatment and retreat, what'sthe prediction and, uh, other
prognostic properties likecancer size, prostate specific
antigen PSA, um, level and a PSAdensity and diagnosis were
(14:50):
evaluated and the results were.
And this particular algorithmhad the specificity of a, um,
specificity, sorry, sensitivityof 0.
96 and specificity 0.
73 respectively.
Um, And their reference methodto check how good it was, was
(15:14):
the original pathology reportdiagnosis.
So that was what they werecomparing against.
Um, and the conclusion is thatthis cancer detection algorithm
could be used to reducepathologists workflow.
So, Could be used, I would haveto go into the, um, original
(15:35):
paper and see, okay, like, arethey planning to use it?
Or did they just have the goodspecificity and sensitivity and
sensitivity and specificity touse it?
I guess, um, but basically itcould be used to reduce
pathologist workflow in theactive surveillance cohort.
Um, And the detected canceramount correlated with the
(15:59):
cancer length measured by thepathologist and the algorithm
performed well in finding evensmall areas of cancer.
Are we like suggesting here thatwe run this algorithm and then,
uh, the pathologist just checksand we have like a screening
tool.
I think that would be cool.
Um, So, and what they say isthat to their knowledge, this is
(16:23):
the first report on an AI basedalgorithm in the world.
An active surveillance cohort.
So normally it's just likediagnostics and but here they
were doing this activesurveillance and it worked.
It worked.
(16:43):
So the 3rd paper today is from,uh, scientific reports and
scientific reports, especially.
Part of the nature portfolio.
So these, when I look at thesepapers, they're usually like
very impactful and prettycomplicated.
Uh, but that's okay.
It means, um, and this group isfrom Wuhan, China, Wuhan, China
(17:12):
published in.
Scientific report.
So the title is matrixmetalloproteinase nine
expression and glioblastomasurvival prediction using
machine learning on digitalpathology images.
So let's start with what arethese matrix metalloproteinases
and why is this expressionimportant?
(17:33):
So these are enzymes that, um,they To change the tumor micro
environment, for example, they,uh, can digest proteins, digest
collagen and restructure,rebuild the tumor micro
environment.
For example, the tumor stromaand, um, what's happening in the
(17:54):
micro environment is oftenimportant for a prognosis and,
but this is an interestingthing.
Set up as well, because theyapplied, so the study aimed to
apply pathomics, I'm like,pathomics, what do you mean with
pathomics?
(18:15):
What is this?
Um, and pathomics is like, Usingpathology expression data and
like multimodal data, includingpathology.
So they'll use this pathoma topredict ma matrix
metalloproteinase expression inglioblastoma and um, to
(18:40):
investigate the underlyingmolecular mechanisms, mechanisms
associated with paths.
So I like heard this word, but Ididn't know it was like a.
Like an official word.
I had to look it up guys.
Um, so they had 120 glioblastomapatients and 78 of them were
(19:01):
allocated to the training, uh,randomly allocated to the
training and test scores forpathomics modeling.
And, um, so then they calculatethe prognostic significance of
this metal, uh, matrixmetalloproteinase.
So just to give you a little bitof context, the normal way of
measuring this is, um, RNA, uh,expression, the RNA levels.
(19:26):
So you have to destroy thetissue and, um.
Find what, what the levels are.
Right?
So here, um, they usedpyrodeomics.
They used pyrodeomics.
I'm going to tell you whatpyrodeomics is because I had no
idea.
I had to look it up for you.
Pyrodeomics was used to extractthe measures of H& E stained
(19:48):
hall slide images.
And I'm like, pyrodeomics?
What is this?
Apparently, it is some opensource model.
Um, I want to read you the realdefinition.
Pyradiomics.
It's an open source model, uh,open source Python package for
(20:09):
the extraction of radiomics datafrom medical images.
I'm like, okay, that's Python.
Pathomics, but they useRadiomics, I guess it worked
well as well.
So anyway, they did featureselection and, um, used
different parameters and theycreated a prediction models, uh,
(20:29):
using support vector machinesand logistic regression.
Um, and, uh, they assess theperformance and what Uh, they
state.
So the performance was assessedusing ROC align analysis,
calibration curve assessment,and decision curve analysis.
(20:52):
Uh, and the MMP nine.
The metal metal, um, metalproteinase nine expression was
elevated in patients withglioblastoma.
So, uh, and this was anindependent prognostic factor
for Glioblastoma.
So this is like independentthing.
(21:13):
And they had other differentfeatures, but, um, I went into
the introduction of this andbasically what they did, they
used the TCGA, uh, Tumor CancerGenome Atlas and another
database, to train this, uh,This model against the RNA
(21:33):
levels, and then they used it,uh, RNA levels for metal protein
is so pretty fancy.
Another, um, thing that gives melike hope.
I don't know if hope is theright word, but basically
another thing where you canpredict something from the image
and the image is always there.
(21:54):
The image is part of yourpathological diagnosis.
So one, one part is okay.
If you can do some imagingwithout, um, the, the slides,
that's fantastic.
That's like the next level.
It would be something that youwill be using for intraoperative
procedures.
Uh, I don't know, maybesomething on the skin where, I
don't know, something where youwould not need to take out the
(22:17):
sample, but for this type ofdiagnosis, the pathology on
glass is often going to be the.
Um, the thing that's going tobe, uh, still going to be used
and this image is always there.
Right?
So if you can avoid somedownstream tests, or if you can
guide or narrow down the teststhat you're going to be doing
(22:39):
with, um, after the pathology isalready done, after the biopsy
is already taken, that's, uh,all power, power to you, right?
Fantastic.
So this is another thing wherethey, where they did that and
published in a journal.
Nature, and these are the threeones that, uh, no, there is one,
(22:59):
one more that I want to sharewith you.
Um, it was interesting because Iwas looking, so, so the title
of, uh, the next one is 1million segmented red blood
cells with.
240k classified in nine shapesand 47, 000 patches of 25 manual
(23:21):
blood smears.
So it's an interesting title,right?
They had 25 manual blood smearsand they managed to have all
those thousands of pieces ofdata.
Um, and, uh, this is also,Scientific data.
I'm checking here.
This is also nature portfolio.
I'm like laughing at this titlebecause it's more like a
marketing title, uh, for peopleto click on.
(23:45):
Um, but you know, that's a goodenough title for scientific
data.
And this group is from Egypt.
And this is actually a groupthat has a commercial entity
pathologics.
I was looking up this company.
If anybody from pathologics, um,Is listening to this, feel free
to reach out to me on linked in,um, in great publication, but
(24:09):
they couldn't find the websiteof this company, but my computer
is acting out.
So what happened here?
I want a different color.
I want a green.
Okay.
So, um, 25, 20 percent ofcomplete blood count samples
(24:30):
necessitate visual review usinglight microscopes, um, or
digital pathology scanners,right?
We can scan as well.
Uh, so there is no currently notechnological alternative to the
visual evaluation of Red bloodcells, which I thought was
interesting because I thoughtthat, um, the blood analysis,
(24:53):
but it's, I think it's, um,white blood cells, like all
these morphologies that, uh, uh,AI algorithms actually started
in cytology to differentiatewhite blood cells and different
blood cells from each other.
So I was surprised to read this,that there is no technological
alternative for red blood cells,morphology shapes.
(25:14):
So, Um, the erythrocytes.
Um, so, um, and, and what arethe problems with erythrocytes?
Um, t True, sorry.
Wrong pen.
There is, um, true non artifactteardrop shaped red blood cells
(25:35):
and schistocytes or fragmentedred blood cells are commonly
associated with serious medicalconditions and can be fatal and
then increased ovalocytes areassociated with almost all type
of anemias.
Interesting.
So.
What they did, they took 25blood smears, um, and each of
(26:01):
those blood smears was from adifferent patient.
Then they were manuallyprepared, stained, and sorted
into four groups, and each groupunderwent treatment.
Imaging using different camerasintegrated into light microscope
with 40 X microscopic lenses.
So they did it with microscopiccamera.
They didn't do this withscanning.
(26:25):
And what happened then was thatthey, uh, had a lot of patches.
So because, uh, with themicroscopic camera, you'll
probably have to go field byfield of view, by field of view,
by field of view.
And, uh, had, they had a lot ofpatches, 47, 000 plus field
images or patches.
(26:45):
And then two.
hematologists processed cell bycell to provide 1, 000, 000 plus
segmented red blood cells withcoordinates and classified 240,
000 of red blood cells into nineshapes.
So we have classification ofnine shapes and this is, um, um,
Manual data labeling, and theycreated a data set of safety
(27:10):
RBCs for AI and safety is one ofthe authors of this paper.
And then you can, you have thisdata set that enables the
development of testing and deeplearning automation for red
blood cells, morphologies andshape examination.
Um, and you can also includespecific normalization for blood
(27:30):
smear stains.
So the blood smears and thecytology, these are slightly
different stains.
Um, then the H and E that we areall used to, all who look at
tissue are used to, um,pathologists a lot.
Um, so they prepared this dataset and they provided also
codes, uh, for one for codes,meaning the code.
(27:56):
I don't know which codinglanguage they used, but one was
for semi automated imageprocessing, and another for
testing of deep learning basedimage classifiers.
So that is interesting.
Guys, which one do you think ismost interesting?
Liver, blood, or What else wehad?
Prostate active surveillanceand, and see, I don't even
(28:22):
remember what we talked aboutand I just finished talking
about it.
And glioblastoma, themetalloproteins.
If you have a favorite, give mea 1, 2, 3, 4.
Um, in the meantime, Thank youso much for joining, but I want
to tell you a few more thingstoday.
So, uh, what's up in general inlife?
I'm going to Poland in August,so I don't have the live stream
(28:47):
scheduled for August yet.
I probably will skip one week.
Um, and.
Have it.
No, I think I'm going to keep itat the same time because in
Poland, it's 6 hours difference.
So it's going to be, uh, 12o'clock my Polish lunchtime.
It's still going to be 6 o'clockfor everybody who is in the same
(29:07):
time zone as I am right now.
So that's going to be happening.
You will have see a differentbackground behind me.
And then another thing I'mworking on.
Let me know if you're interestedin this, uh, it's gonna be, I
have a bunch of YouTube videos,as you know, um, it's like.
(29:28):
Almost 370 or over 400.
And so anytime there was a topicI want to explain, I would make
a YouTube video and one of mymembership digital pathology
club members suggested, Hey,could you like embed those
videos in a kind of curriculumand, um, I am looking for a way
(29:53):
to choose, uh, my YouTubevideos.
So, um, this is, you know,everything is already available
for free, but I want to have alittle small, um, Option like a
paid version where they aresystematized in the curriculum,
um, that's going to be somethingthat is less, uh, investment
(30:13):
than my digital pathology clubmembership, which is at 97 a
month.
This is going to be like a onetime course fee.
So if this is somethinginteresting for you.
Give me a comment, say YouTubecourse, and I'm going to gauge
interest.
How many of you are interestedin that in this?
It's like, I call it minicourse, but it's not going to be
(30:35):
mini course.
It's going to be likecomprehensive curriculum, but
from videos that are alreadyavailable for free, that are
going to be systematized, maybesome additional resources.
And I will need to host itsomewhere so that it's not, uh,
I mean, the videos are going tobe the same, but I'm going to
host it somewhere where youdon't have to scroll through
YouTube, It's going to be adedicated page.
(30:59):
So if this is of any interest,drop a comment, YouTube course.
And another cool thing, uh, thatI witnessed or, or was part of,
I was just listening to it.
There was an AWS, Amazon, um,works services.
What, what does AWS stand for?
Even I know it's Amazon forcloud, right?
(31:22):
AWS.
Um, What is it?
Amazon web services.
Yeah.
Amazon for cloud web services.
Um, they had a summit in NewYork and, uh, there was a part
of the summit where they wereannouncing different
collaborations with differentpeople, different businesses
(31:42):
that you could register for andlisten to.
So.
There's a lot going on withgenerative AI, right?
So, by the way, if generative AIand chat GPT and stuff like that
is of interest, and drop me acomment, uh, below as well, that
this is a topic of interest, andthen I'm going to go Um, into,
undermine the AI and do aspecific search for that.
(32:06):
And then we're gonna go throughpapers on that particular topic.
Um, I remember giving a webinarlast year where, uh, there were
no, uh, I was explaining chat,GPT and you can find it on
YouTube as well.
Maybe I'm gonna link, uh, in thecomment.
But basically there were like, Iwas doing literature research
and there was nothing out.
And now every week.
Um, there is something new, uh,pretty high publication.
(32:30):
Uh, it was tested howpathologists would like to use
it.
Um, so, um, if that's a topic ofinterest, give me a comment.
Let me know chat GPT orgenerative AI, uh, whatever you
want to put in there.
Uh, so, uh, what's happening,they are working, AWS is working
(32:51):
with Anthropic.
Anthropic is like, uh, I don'tknow if they're, I assume
they're competitor to OpenAI.
They have this other model,Claude, in contrast to ChatGPT,
and I use both.
By the way, uh, so they areworking with Claude, uh, they're
working on solutions forenterprises that are, um, secure
(33:11):
that are, uh, being able to befine tuned, uh, and, uh, just
use your secure company datawithout leaking everywhere.
So a lot of, um.
Business solutions that willhelp different type of
businesses, um, get leverage.
And I, of course, was on thelookout for anything in the
biotech space or pharma orpathology.
(33:34):
Um, and I've heard that Pfizeris already using using Anthropic
for something.
I was looking for some pressreleases and it only says that
they are collaborating anddoesn't really say, uh, what.
Um.
What they're doing with it.
I couldn't figure that out.
And okay, Erica, perfect.
You want the chat GPT version.
(33:56):
If anybody else, uh, let me knowbecause then I can prepare
something super specific.
And if you have any other topicsthat you would be interested in,
rather than just the abstractreviews, let me know in the
comments and in the meantime,you have a wonderful rest of
your day and I talk to you inthe next episode..
(34:18):
Just very quickly because youare listening to it.
I just wanted to let you knowthat everything that I'm like
asking in the comments oreverything, you can just send it
to me by email.
If you're subscribed to mynewsletter, you're gonna get,
you're getting.
All the content that is goingout there.
And you can always respond tothose emails and I am super,
(34:40):
super happy to hear from youanytime.
And if you like this show, thereis also an option to support the
show.
Either on the podcast website.
Or if you're watching this onYouTube, there is an option.
To give super thanks.
And if you are listening to thispodcast, because you want to
hear interviews with people,don't worry.
(35:00):
There are going to be interviewsnext week.
We're releasing another episode.
With a digital pathologist, Dr.
Todd Randolph.
So stay tuned.
We're going to have both anamazing guest and we're going to
be reviewing abstracts to standtop off that digital pathology,
AI science.
Welcome my digital pathologytrailblazers.
Today was the third digit pathdigest abstract review.
And my friends, this is becominga thing.
I see people who are joiningover and over again.
And.
It's at 6:00 AM in the us.
So I know that people, well, no,I take the back because I had
(00:23):
one person who, uh, is joining.
I think he's he joined.
I already like the third door,the second time.
Anyway, so there are people inthe us.
Cool.
The morning.
So thank you so much.
And when I was reviewing theabstracts today, there is a
theme that these publicationsare not really from pathology
journals.
So I'm not going to be givingyou more intro because there's
(00:43):
intro already in the digitaldigest.
So let's dive into it.
Learn about the newest digitalpathology trends in science and
industry.
Meet the most interesting peoplein the niche, and gain insights
relevant to your own projects.
Here is where pathology meetscomputer science.
(01:04):
You are listening to the DigitalPathology Podcast with your
host, Dr.
Aleksandr Zhurav.
Aleksandra (01:14):
Good morning.
Good morning, everyone.
It's 6 a.
m.
in Pennsylvania, and I welcomeyou to the third edition of
DigiPath Digest.
When you log in here, let meknow when you are tuning in
(01:34):
from.
Give me a hi and when Whereyou're where you're at right
now.
So, um, let me tell you a story.
Of course.
I have my beautiful, fantasticnew camera here.
That didn't work.
So I'm just doing my normal, um,computer camera.
But let me know where you'retuning in from.
(01:57):
I see you're joining and let mejust say hello in the comments
and welcome you.
Um, we have a few abstractstoday.
Three.
Oh, Erica.
So great to have you here.
Hello.
(02:18):
And Robert, you made it fromMaine.
Yeah.
So there was a little bit of aglitch yesterday.
I think I had it initially foryesterday, um, at 6.
00 PM.
And, uh, it's always at 6 AM inthe morning, but, um, Robert
sent me a message that, Hey, isit yesterday or today it's
(02:39):
today.
And we have Jason fromBaltimore.
Baltimore is not that far fromme.
It's one and a half hours.
So fantastic to have you hereguys.
Um, I know there's going to bemore people joining, but for
those who already were here ontime, um, here agenda guys, I'm
serious.
Um, so.
For those who are here, butdon't know me, I'm Dr.
(03:00):
Alex Rudaf and I'm a veterinarypathologist and the founder of
Digital Pathology Place.
And I'm on a mission to teachyou all you need to know about
digital pathology.
And part of this mission is mybook.
If you didn't grab the book yet,there is one when you go to my
website, digitalpathologyplace.
com, you can download this onefor free and Let's dive into one
(03:25):
question before, um, because,uh, so now what I'm doing, I'm
reviewing the abstracts thatcome from PubMed alerts, but I
also, uh, discovered this, um,software for literature review,
literature research calledUndermined and I showed it to
(03:45):
you last time, but I'm going toshare it with you one more time
because I want to ask yousomething.
Step screen, unsure screen, uh,so undermined, you should see it
now.
So this is this, uh, researchengine.
(04:06):
So, um, what, um, last, thisweek I had a, uh, ergonomic
training and.
Welcome.
Ergonomics training at work.
And, uh, I was like, okay, I'mworking remotely.
So I do have some stuff.
I have like a standing desk.
I have a walking treadmill forthings because when I started
(04:30):
working from home and I have thewatch and I have the ring to
measure my activity, theactivity goes, uh, a lot.
Like it's so much less than whenI was actually driving to work
and I had to go from my car tosomewhere anyway.
And, uh, welcome Milena from theNetherlands.
Great to have you here anyway.
So, uh, long story short withthis tool undermined, you can do
(04:52):
research on specific topic.
And I did a research on, um,pathology, digital pathology
ergonomics.
So if this is something you'reinterested in me covering, I
mean, I'm going to cover this,um, At some point anyway, but if
you're interested in any othertopics, leave me a comment, what
topics you're interested in,like specific topics, like, I
(05:14):
don't know the scanners or, um,the AI models that are
available, uh, open source orsomething, whatever, uh, you're
interested in and hi to India.
And welcome, Ahmed from SaudiArabia.
Great to have you here, guys.
(05:34):
So any topic that you think, oh,I would need, like, I would want
something specific to that.
Not just what comes from PubMedevery week, but just specific to
that particular topic.
Let me know in the comments, thetopic.
And without further ado, let'smove to our abstracts.
(05:58):
Need to share a differentscreen, which is my tablet where
I can also draw, right?
Yes.
Okay.
Perfect.
So today the first topic is AIbased digital pathology provides
(06:19):
newer insights into lifestyleintervention induced fibrosis
regression in MASLD.
Thank you.
So, um, this is liver fibrosis.
I'm going to explain everythingto you, but, um, something
that's interesting, I'm lookingwhere these publications come
from, not only geographically.
So, um, this one is from China,but what kind of journals, uh,
(06:44):
they are in.
So this is like a liver internalmedicine journal.
Um, there are a few that wedon't have, uh, uh, In pathology
journals today, and they arefrom other disciplines where
people use digital pathologyand, and published there, which
I think is cool because justlike spreads digital pathology
from this niche of pathologists,because pathology is the gateway
(07:06):
to diagnosis.
And, um, that's why I'm happyabout that.
So.
What did they say here today?
Life stay intervention is themain stay of therapy for
metabolic dysfunctionassociated, uh, stato hepatitis.
So MASH, this is what we'retalking about.
(07:28):
And fibrosis is a keyconsequence that predicts
adverse clinical outcomes.
So when it goes very bad, youhave liver fibrosis.
I want to make it even bigger sothat you can see this.
Um, anyway, so what they didhere, they did imaging of
biopsies of pre treatment andpost treatment biopsies.
(07:48):
And I'm going to tell you whatthe treatment or like, um, what
the intervention was, um, butthey use a different method than
just our standard H and E.
They used harmonic generation tophoton excitation fluorescence
microscopy with, um, AI imageanalysis.
So, uh, it's, uh.
(08:08):
non staining method.
You excite the surface of thesample and you image it, you
image the fluorescence and themethod, um, to calculate the
level of fibrosis.
They called it Q fibrosis.
And this is, this providesqualitative assessment.
And so I thought that was alsointeresting because the non
(08:30):
staining methods are gettingmore popular.
prevalent.
And if you're just, if you'vejust joined because I see the
number of people joining andunjoining fluctuate, uh, be sure
to say hi and be sure to let meknow if there is a specific
topic you would like to cover inone of these, uh, journal clubs.
So, um, they examined thoseunstained sections.
(08:55):
So here's the keyword unstainedsections from paired liver
biopsies, baseline and end oftreatment intervention.
So pre and post, and they hadsomething that is called routine
lifestyle intervention andstrengthened lifestyle
intervention.
Um, And they actually sawfibrosis regression, which I
(09:16):
think is a big deal becausefibrosis is like, um, one of
those processes in pathologythat doesn't go away that
easily, but they saw it withthis lifestyle intervention.
Uh, and.
They basically say that usingdigital pathology, they could
detect more pronounced fibrosisregression.
(09:37):
Um, congratulations to them.
Let's see.
Where is it published?
And we already checked in theliver internal medicine.
Okay.
So that is our first abstractwithout staining.
I think this is cool because,um, so I told you the other day,
(09:59):
or I keep telling people, uh,that I was super excited about
the molecular predictions.
And for a long time, like 34years, uh, it was like only
academia, only research, nothinginto the clinic until there was
a publication in 2023, thatsomething actually made it in
(10:20):
the clinic.
The other thing that I'm superexcited about is the stainless
steel.
Uh, imaging direct to digitalimaging.
Sometimes you can do direct todigital from, um, from tissue,
like immediately without evenfixating the tissue or just like
cut it or freeze, cut it ordifferent methods.
(10:42):
And sometimes you need to fixit, but you don't have to stain
it.
So anyway, in each of these.
Uh, options, you skip some ofthe analog stacks, um, analog
steps of pathology of digitalpathology.
You go direct to digital and,and that's something that is
super, um, like a huge advantageof digital pathology because
(11:04):
radiology and all the otherdigital imaging, they have it.
They don't have to go to analogand then to digital like
pathology.
And I'm like, okay, until thatgoes away, uh, we're not going
to be as good as the other ones.
So Uh, we are as good as theother ones, uh, but just have
class slides.
Okay.
Now moving on to our nextabstract, uh, which is
(11:29):
artificial intelligence for, oh,sorry.
Artificial intelligence fordetection of prostate cancer in
biopsies during activesurveillance.
So, um, I assumed that Malepeople know what that is, but I
checked like, what's thedifference between active
surveys and surveillance andtreatment and basically active
(11:54):
surveillance is when you are nottreating, you're just monitoring
and you're checking.
You still do.
Um, biopsies pre and post atsome point, um, sorry, no, you
do the 1st biopsy and then youhave the results that are
telling you that there is nocancer or inconclusive for like,
low.
There is a specific Gleasonscore for that.
(12:15):
And, um.
That's active surveillance.
You don't do prostatectomy, youdon't treat, you just monitor.
And what happened here, they usean AI algorithm, which to me, it
was like, okay, um, well, weeven have an AI algorithm that
is, uh, cleared by the FDA forprostate.
(12:36):
So that's kind of like not thatnovel.
Let me check the journal again.
Because this one was also, Oh,so I didn't know this journal
BJU.
This is a British journal ofurology.
So this is again, a nonpathology journal that, um, is
(12:57):
actually publishing aboutdigital pathology and has the
keywords digital pathology.
So Um, they evaluated cancerdetect, uh, sorry to, uh, uh,
the objectives were to evaluatea cancer detecting artificial
intelligence algorithm on activesurveillance.
So they had, uh, and this is agroup from Norway, they had 180
(13:18):
patients in the, um, prostatecancer research, international
active surveillance priors.
Did I tell you that thepathology likes those, uh,
abbreviations?
Um, I think last time I toldyou.
So, uh, they had diagnostic.
So that the initial diagnosis,initial biopsy and let me zoom
(13:41):
in and their re biopsy slide.
So they had 2 biopsies and thedates were from 2011 to 2000.
20.
And they had a lot of thesebiopsies because they had 4,
000, um, 4, 744.
And they were scanned andanalyzed by an in house AI
(14:04):
cancer, uh, detection team.
Algorithm.
So they, they developed theirown, they didn't use any
commercial tool.
They just developed their own.
And what happened, uh, also theywere, uh, then, um, the goal
was, uh, that To analyze this,um, this algorithm for a
(14:25):
sensitivity specificity andaccuracy to predict the need for
active treatment.
So their surveillance and ifthere is a need for active
treatment and retreat, what'sthe prediction and, uh, other
prognostic properties likecancer size, prostate specific
antigen PSA, um, level and a PSAdensity and diagnosis were
(14:50):
evaluated and the results were.
And this particular algorithmhad the specificity of a, um,
specificity, sorry, sensitivityof 0.
96 and specificity 0.
73 respectively.
Um, And their reference methodto check how good it was, was
(15:14):
the original pathology reportdiagnosis.
So that was what they werecomparing against.
Um, and the conclusion is thatthis cancer detection algorithm
could be used to reducepathologists workflow.
So, Could be used, I would haveto go into the, um, original
(15:35):
paper and see, okay, like, arethey planning to use it?
Or did they just have the goodspecificity and sensitivity and
sensitivity and specificity touse it?
I guess, um, but basically itcould be used to reduce
pathologist workflow in theactive surveillance cohort.
Um, And the detected canceramount correlated with the
(15:59):
cancer length measured by thepathologist and the algorithm
performed well in finding evensmall areas of cancer.
Are we like suggesting here thatwe run this algorithm and then,
uh, the pathologist just checksand we have like a screening
tool.
I think that would be cool.
Um, So, and what they say isthat to their knowledge, this is
(16:23):
the first report on an AI basedalgorithm in the world.
An active surveillance cohort.
So normally it's just likediagnostics and but here they
were doing this activesurveillance and it worked.
It worked.
(16:43):
So the 3rd paper today is from,uh, scientific reports and
scientific reports, especially.
Part of the nature portfolio.
So these, when I look at thesepapers, they're usually like
very impactful and prettycomplicated.
Uh, but that's okay.
It means, um, and this group isfrom Wuhan, China, Wuhan, China
(17:12):
published in.
Scientific report.
So the title is matrixmetalloproteinase nine
expression and glioblastomasurvival prediction using
machine learning on digitalpathology images.
So let's start with what arethese matrix metalloproteinases
and why is this expressionimportant?
(17:33):
So these are enzymes that, um,they To change the tumor micro
environment, for example, they,uh, can digest proteins, digest
collagen and restructure,rebuild the tumor micro
environment.
For example, the tumor stromaand, um, what's happening in the
(17:54):
micro environment is oftenimportant for a prognosis and,
but this is an interestingthing.
Set up as well, because theyapplied, so the study aimed to
apply pathomics, I'm like,pathomics, what do you mean with
pathomics?
(18:15):
What is this?
Um, and pathomics is like, Usingpathology expression data and
like multimodal data, includingpathology.
So they'll use this pathoma topredict ma matrix
metalloproteinase expression inglioblastoma and um, to
(18:40):
investigate the underlyingmolecular mechanisms, mechanisms
associated with paths.
So I like heard this word, but Ididn't know it was like a.
Like an official word.
I had to look it up guys.
Um, so they had 120 glioblastomapatients and 78 of them were
(19:01):
allocated to the training, uh,randomly allocated to the
training and test scores forpathomics modeling.
And, um, so then they calculatethe prognostic significance of
this metal, uh, matrixmetalloproteinase.
So just to give you a little bitof context, the normal way of
measuring this is, um, RNA, uh,expression, the RNA levels.
(19:26):
So you have to destroy thetissue and, um.
Find what, what the levels are.
Right?
So here, um, they usedpyrodeomics.
They used pyrodeomics.
I'm going to tell you whatpyrodeomics is because I had no
idea.
I had to look it up for you.
Pyrodeomics was used to extractthe measures of H& E stained
(19:48):
hall slide images.
And I'm like, pyrodeomics?
What is this?
Apparently, it is some opensource model.
Um, I want to read you the realdefinition.
Pyradiomics.
It's an open source model, uh,open source Python package for
(20:09):
the extraction of radiomics datafrom medical images.
I'm like, okay, that's Python.
Pathomics, but they useRadiomics, I guess it worked
well as well.
So anyway, they did featureselection and, um, used
different parameters and theycreated a prediction models, uh,
(20:29):
using support vector machinesand logistic regression.
Um, and, uh, they assess theperformance and what Uh, they
state.
So the performance was assessedusing ROC align analysis,
calibration curve assessment,and decision curve analysis.
(20:52):
Uh, and the MMP nine.
The metal metal, um, metalproteinase nine expression was
elevated in patients withglioblastoma.
So, uh, and this was anindependent prognostic factor
for Glioblastoma.
So this is like independentthing.
(21:13):
And they had other differentfeatures, but, um, I went into
the introduction of this andbasically what they did, they
used the TCGA, uh, Tumor CancerGenome Atlas and another
database, to train this, uh,This model against the RNA
(21:33):
levels, and then they used it,uh, RNA levels for metal protein
is so pretty fancy.
Another, um, thing that gives melike hope.
I don't know if hope is theright word, but basically
another thing where you canpredict something from the image
and the image is always there.
(21:54):
The image is part of yourpathological diagnosis.
So one, one part is okay.
If you can do some imagingwithout, um, the, the slides,
that's fantastic.
That's like the next level.
It would be something that youwill be using for intraoperative
procedures.
Uh, I don't know, maybesomething on the skin where, I
don't know, something where youwould not need to take out the
(22:17):
sample, but for this type ofdiagnosis, the pathology on
glass is often going to be the.
Um, the thing that's going tobe, uh, still going to be used
and this image is always there.
Right?
So if you can avoid somedownstream tests, or if you can
guide or narrow down the teststhat you're going to be doing
(22:39):
with, um, after the pathology isalready done, after the biopsy
is already taken, that's, uh,all power, power to you, right?
Fantastic.
So this is another thing wherethey, where they did that and
published in a journal.
Nature, and these are the threeones that, uh, no, there is one,
(22:59):
one more that I want to sharewith you.
Um, it was interesting because Iwas looking, so, so the title
of, uh, the next one is 1million segmented red blood
cells with.
240k classified in nine shapesand 47, 000 patches of 25 manual
(23:21):
blood smears.
So it's an interesting title,right?
They had 25 manual blood smearsand they managed to have all
those thousands of pieces ofdata.
Um, and, uh, this is also,Scientific data.
I'm checking here.
This is also nature portfolio.
I'm like laughing at this titlebecause it's more like a
marketing title, uh, for peopleto click on.
(23:45):
Um, but you know, that's a goodenough title for scientific
data.
And this group is from Egypt.
And this is actually a groupthat has a commercial entity
pathologics.
I was looking up this company.
If anybody from pathologics, um,Is listening to this, feel free
to reach out to me on linked in,um, in great publication, but
(24:09):
they couldn't find the websiteof this company, but my computer
is acting out.
So what happened here?
I want a different color.
I want a green.
Okay.
So, um, 25, 20 percent ofcomplete blood count samples
(24:30):
necessitate visual review usinglight microscopes, um, or
digital pathology scanners,right?
We can scan as well.
Uh, so there is no currently notechnological alternative to the
visual evaluation of Red bloodcells, which I thought was
interesting because I thoughtthat, um, the blood analysis,
(24:53):
but it's, I think it's, um,white blood cells, like all
these morphologies that, uh, uh,AI algorithms actually started
in cytology to differentiatewhite blood cells and different
blood cells from each other.
So I was surprised to read this,that there is no technological
alternative for red blood cells,morphology shapes.
(25:14):
So, Um, the erythrocytes.
Um, so, um, and, and what arethe problems with erythrocytes?
Um, t True, sorry.
Wrong pen.
There is, um, true non artifactteardrop shaped red blood cells
(25:35):
and schistocytes or fragmentedred blood cells are commonly
associated with serious medicalconditions and can be fatal and
then increased ovalocytes areassociated with almost all type
of anemias.
Interesting.
So.
What they did, they took 25blood smears, um, and each of
(26:01):
those blood smears was from adifferent patient.
Then they were manuallyprepared, stained, and sorted
into four groups, and each groupunderwent treatment.
Imaging using different camerasintegrated into light microscope
with 40 X microscopic lenses.
So they did it with microscopiccamera.
They didn't do this withscanning.
(26:25):
And what happened then was thatthey, uh, had a lot of patches.
So because, uh, with themicroscopic camera, you'll
probably have to go field byfield of view, by field of view,
by field of view.
And, uh, had, they had a lot ofpatches, 47, 000 plus field
images or patches.
(26:45):
And then two.
hematologists processed cell bycell to provide 1, 000, 000 plus
segmented red blood cells withcoordinates and classified 240,
000 of red blood cells into nineshapes.
So we have classification ofnine shapes and this is, um, um,
Manual data labeling, and theycreated a data set of safety
(27:10):
RBCs for AI and safety is one ofthe authors of this paper.
And then you can, you have thisdata set that enables the
development of testing and deeplearning automation for red
blood cells, morphologies andshape examination.
Um, and you can also includespecific normalization for blood
(27:30):
smear stains.
So the blood smears and thecytology, these are slightly
different stains.
Um, then the H and E that we areall used to, all who look at
tissue are used to, um,pathologists a lot.
Um, so they prepared this dataset and they provided also
codes, uh, for one for codes,meaning the code.
(27:56):
I don't know which codinglanguage they used, but one was
for semi automated imageprocessing, and another for
testing of deep learning basedimage classifiers.
So that is interesting.
Guys, which one do you think ismost interesting?
Liver, blood, or What else wehad?
Prostate active surveillanceand, and see, I don't even
(28:22):
remember what we talked aboutand I just finished talking
about it.
And glioblastoma, themetalloproteins.
If you have a favorite, give mea 1, 2, 3, 4.
Um, in the meantime, Thank youso much for joining, but I want
to tell you a few more thingstoday.
So, uh, what's up in general inlife?
I'm going to Poland in August,so I don't have the live stream
(28:47):
scheduled for August yet.
I probably will skip one week.
Um, and.
Have it.
No, I think I'm going to keep itat the same time because in
Poland, it's 6 hours difference.
So it's going to be, uh, 12o'clock my Polish lunchtime.
It's still going to be 6 o'clockfor everybody who is in the same
(29:07):
time zone as I am right now.
So that's going to be happening.
You will have see a differentbackground behind me.
And then another thing I'mworking on.
Let me know if you're interestedin this, uh, it's gonna be, I
have a bunch of YouTube videos,as you know, um, it's like.
(29:28):
Almost 370 or over 400.
And so anytime there was a topicI want to explain, I would make
a YouTube video and one of mymembership digital pathology
club members suggested, Hey,could you like embed those
videos in a kind of curriculumand, um, I am looking for a way
(29:53):
to choose, uh, my YouTubevideos.
So, um, this is, you know,everything is already available
for free, but I want to have alittle small, um, Option like a
paid version where they aresystematized in the curriculum,
um, that's going to be somethingthat is less, uh, investment
(30:13):
than my digital pathology clubmembership, which is at 97 a
month.
This is going to be like a onetime course fee.
So if this is somethinginteresting for you.
Give me a comment, say YouTubecourse, and I'm going to gauge
interest.
How many of you are interestedin that in this?
It's like, I call it minicourse, but it's not going to be
(30:35):
mini course.
It's going to be likecomprehensive curriculum, but
from videos that are alreadyavailable for free, that are
going to be systematized, maybesome additional resources.
And I will need to host itsomewhere so that it's not, uh,
I mean, the videos are going tobe the same, but I'm going to
host it somewhere where youdon't have to scroll through
YouTube, It's going to be adedicated page.
(30:59):
So if this is of any interest,drop a comment, YouTube course.
And another cool thing, uh, thatI witnessed or, or was part of,
I was just listening to it.
There was an AWS, Amazon, um,works services.
What, what does AWS stand for?
Even I know it's Amazon forcloud, right?
(31:22):
AWS.
Um, What is it?
Amazon web services.
Yeah.
Amazon for cloud web services.
Um, they had a summit in NewYork and, uh, there was a part
of the summit where they wereannouncing different
collaborations with differentpeople, different businesses
(31:42):
that you could register for andlisten to.
So.
There's a lot going on withgenerative AI, right?
So, by the way, if generative AIand chat GPT and stuff like that
is of interest, and drop me acomment, uh, below as well, that
this is a topic of interest, andthen I'm going to go Um, into,
undermine the AI and do aspecific search for that.
(32:06):
And then we're gonna go throughpapers on that particular topic.
Um, I remember giving a webinarlast year where, uh, there were
no, uh, I was explaining chat,GPT and you can find it on
YouTube as well.
Maybe I'm gonna link, uh, in thecomment.
But basically there were like, Iwas doing literature research
and there was nothing out.
And now every week.
Um, there is something new, uh,pretty high publication.
(32:30):
Uh, it was tested howpathologists would like to use
it.
Um, so, um, if that's a topic ofinterest, give me a comment.
Let me know chat GPT orgenerative AI, uh, whatever you
want to put in there.
Uh, so, uh, what's happening,they are working, AWS is working
(32:51):
with Anthropic.
Anthropic is like, uh, I don'tknow if they're, I assume
they're competitor to OpenAI.
They have this other model,Claude, in contrast to ChatGPT,
and I use both.
By the way, uh, so they areworking with Claude, uh, they're
working on solutions forenterprises that are, um, secure
(33:11):
that are, uh, being able to befine tuned, uh, and, uh, just
use your secure company datawithout leaking everywhere.
So a lot of, um.
Business solutions that willhelp different type of
businesses, um, get leverage.
And I, of course, was on thelookout for anything in the
biotech space or pharma orpathology.
(33:34):
Um, and I've heard that Pfizeris already using using Anthropic
for something.
I was looking for some pressreleases and it only says that
they are collaborating anddoesn't really say, uh, what.
Um.
What they're doing with it.
I couldn't figure that out.
And okay, Erica, perfect.
You want the chat GPT version.
(33:56):
If anybody else, uh, let me knowbecause then I can prepare
something super specific.
And if you have any other topicsthat you would be interested in,
rather than just the abstractreviews, let me know in the
comments and in the meantime,you have a wonderful rest of
your day and I talk to you inthe next episode..
(34:18):
Just very quickly because youare listening to it.
I just wanted to let you knowthat everything that I'm like
asking in the comments oreverything, you can just send it
to me by email.
If you're subscribed to mynewsletter, you're gonna get,
you're getting.
All the content that is goingout there.
And you can always respond tothose emails and I am super,
(34:40):
super happy to hear from youanytime.
And if you like this show, thereis also an option to support the
show.
Either on the podcast website.
Or if you're watching this onYouTube, there is an option.
To give super thanks.
And if you are listening to thispodcast, because you want to
hear interviews with people,don't worry.
(35:00):
There are going to be interviewsnext week.
We're releasing another episode.
With a digital pathologist, Dr.
Todd Randolph.
So stay tuned.
We're going to have both anamazing guest and we're going to
be reviewing abstracts to standtop off that digital pathology,
AI science.
Popular Podcasts
Are You A Charlotte?
In 1997, actress Kristin Davis’ life was forever changed when she took on the role of Charlotte York in Sex and the City. As we watched Carrie, Samantha, Miranda and Charlotte navigate relationships in NYC, the show helped push once unacceptable conversation topics out of the shadows and altered the narrative around women and sex. We all saw ourselves in them as they searched for fulfillment in life, sex and friendships. Now, Kristin Davis wants to connect with you, the fans, and share untold stories and all the behind the scenes. Together, with Kristin and special guests, what will begin with Sex and the City will evolve into talks about themes that are still so relevant today. "Are you a Charlotte?" is much more than just rewatching this beloved show, it brings the past and the present together as we talk with heart, humor and of course some optimism.
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