TPES 2024: Dr Fabio Papa "Image misinterpretation"
Watch on YouTubeVideo summary
Dr. Fabio Papa, a distinguished staff anesthesiologist and fellowship director at St. Michael's Hospital in Toronto, delivers a critical lecture on the pervasive issue of image misinterpretation during echocardiography. His presentation focuses primarily on imaging artifacts—display phenomena that do not accurately represent anatomical structures—and categorizes them into four main types: false positives where non-existent structures appear present, false negatives where real structures are hidden or absent, distorted representations where existing structures look different from reality, and misplaced images showing structures in incorrect locations. Dr. Papa explains that these artifacts arise fundamentally from the physics of ultrasound waves interacting with tissues; specifically, how sound waves reflect off boundaries between materials with different acoustic impedances and refract at angles determined by those properties. Crucially, he notes that because an ultrasound machine cannot directly observe what happens to a wave after it is emitted until it returns as an echo, the software must make four key assumptions: that waves travel in straight lines, structures are imaged only once, all intervening structures generate reflections, and image position correlates with travel time. When any of these assumptions are violated due to tissue interfaces or equipment limitations, artifacts inevitably occur.
The speaker details several specific mechanisms behind common pitfalls, starting with reverberation artifacts, which happen when sound waves bounce back and forth between two strong reflectors like the ascending aorta and pulmonary artery, creating ghost images at double the depth of the original structure. He distinguishes this from other variations such as comet-tail or ring-down artifacts caused by air bubbles near prosthetic valves. Another major category is acoustic shadowing, where dense structures like mechanical heart valves or calcified tissues block sound waves entirely, obscuring deeper anatomy and potentially hiding pathology behind them. Refraction can cause duplication of images, while mirror imaging creates false duplicates across strong reflectors like the chest wall. Additionally, side-lobe artifacts occur when weaker lateral beams hit a strong reflector and are misinterpreted by the system as coming from the main beam's center line, often mimicking dissections or masses in distant areas. In three-dimensional echocardiography, unique issues arise such as stitching artifacts that make suture lines appear thicker than they are due to resolution differences between planes, blooming effects around metallic devices, and dropout artifacts where parts of structures like the aortic valve disappear during diastole but reappear on 2D views.
To mitigate these risks, Dr. Papa emphasizes practical strategies including changing the angle of incidence, altering probe orientation, adjusting ultrasound settings, and most importantly, correlating findings with clinical history before making diagnostic decisions. He illustrates this necessity through several case studies where initial misinterpretations nearly led to incorrect surgical interventions or cancellations. In one instance, a large pericardial fat pad was mistaken for an effusion compressing the right ventricle in an unstable patient; recognizing it as normal anatomical variation prevented unnecessary surgery and allowed clinical management instead. Another case involved massive gastric fluid mimicking a dissection flap near the heart arch until decompression clarified the image, while yet another demonstrated how regurgitant flow across a prosthetic valve could be confused with mitral regurgitation or thrombus if not carefully analyzed without color Doppler. He also highlights rare scenarios like left atrial appendage invagination appearing as new clots during bypass weaning and retained surgical holders mimicking severe tricuspid stenosis years after replacement, underscoring that many "pathologies" are actually normal variants or technical artifacts known to experienced practitioners but dangerous for those without a high index of suspicion.
Ultimately, the core message is that avoiding diagnostic traps requires a disciplined approach involving multiple views and modalities whenever an image seems suspicious or inconsistent with clinical findings. Dr. Papa warns against relying on single images, especially in complex cases involving cardiac devices or difficult acoustic windows, as errors can lead to vastly different patient outcomes ranging from unnecessary surgeries to delayed treatments for actual conditions like thrombi. By understanding the underlying physics of ultrasound and recognizing the specific signatures of common artifacts such as reverberations, shadows, refractions, and side lobes, clinicians can differentiate between true pathology and misleading visual illusions. His concluding take-home points reinforce that stopping to think critically about every unexpected finding, verifying it across different planes, and maintaining awareness of how devices alter image quality are essential skills for any echocardiographer aiming to provide accurate diagnoses in daily practice.
Read the full video transcript
it now gives me a great pleasure to
introduce Dr Fabio Papa Dr Papa
graduated from the University of sa
Paulo at reoo school of medicine where
he also completed residency he then
pursued a clinical Fellowship in
cardiovascular anesthesia at the
University of Toronto he obtained a MERS
in health practitioner in clinical
education at the dalana school of Public
Health at the University of Toronto he
is a fellow of the American Society of
echocardiography throughout his career
Dr Papa has been actively involved in
teaching and education peroperative
trans soft echocardiography and point of
care ultrasound also recently joining us
in Victoria for our annual anesthesia
conference there he has been both
locally and internationally involved in
curriculum development he's currently a
staff anesthesiologist at St Michael's
Hospital in Toronto where he holds the
position of assistant professor at the
University of Toronto and he's also the
cardiovascular anesthesia Fellowship
director
at St Michael's Hospital Fabio thank you
very much for joining us we're really
looking forward to your lecture and I'll
hand the floor over to you thank you
Amari thank you for the kind um
introduction thank you for you and
Marcos for the invitation it's a
pleasure to be here um it's not the
first year and but it's always a
pleasure to be participating in this
conference so hi everyone the topic of
my presentation is image
misinterpretation so I have no
disclosures and over the next 20 minutes
or so by no means um I'll go over um
comprehensive evaluation and discussion
about um Imaging artifacts but I'm going
to review some of the most common
Imaging artifacts that we encounter in
our clinical practice discussing ways to
help to avoid or at least identify um
Imaging uh
misinterpretations the last part of the
presentation I'm going to use some
examples that encountered in our daily
practice uh in order to try to
differentiate through pathologies from
misinterpretations what is um Imaging
artifact I like this uh definition is
simple it's a display phenomena that do
not properly represent the imagy
structures and there are four different
classifications of Imaging artifacts so
uh the structures they appear to be
there and in fact they are not one one
example that I like to mention every
time I I think about this is
reverberations in the ascending aorta
and sometimes mimicking a typ a
Dissection the second typ type is when
structures are not there and they they
do not appear to be there when in fact
they are I think this one is easy to
examplify uh thinking about um shadowing
caused by the mitro valve when you
cannot see the LV
properly sometimes the structures can
look differently from reality and I'm
going to when I think about this I think
about blurring and blooming artifact in
the micro valve on 3G looking from the
left AUM which I'm going to um describe
a little bit later in the presentation
and also any structures that appear to
be in the wrong location especially now
when you do a lot of um structural hard
procedures with plater devices and um
and uh devices also in the left
appendage normally the Imaging artifacts
they are normally they occurs for two
different reasons the most common is
related to how the ultrasound wave when
traveling from one um um from one
instructure to the other how the
ultrasound wave behaves in terms of
reflection or refraction and it's also
um another source of um Imaging artifact
is how um is related to the ultrasound
be properties we can Al we cannot uh
forget also the Pres
of cardiac devices that can also cause
artifacts and also you cannot forget the
presence of artifacts not only in 2G but
also in the 3G
modality two two special considerations
we need to we need to make when thinking
about um Imaging artifacts the the first
one is how the ultrasound wave travels U
between two different tissues for
example we have the ultrasound probe the
ultrasound probe emits uh the first the
ultrasound wave the ultrasound wave
travels through the tissue one which has
um a different um acoustic imped from
the tissue number two when the
ultrasound wave hits the boundary
between these two tissues what happens
with the ultrasound wave are twofold the
ultrasound wave can be reflected most of
the time with the same angle of um of um
um the same angle that was um U used for
the for the income wave
or the ultrasound wave can be refracted
um most of the time with different angle
because the two tissues they have
different AC acoustic
impedances the other thing that is
important to mention is that like from
the moment that the ultrasound wave is
uh emitted by the ultrasound probe until
the moment it the ultrasound probe
receives the ultrasound wave the
ultrasound wave is pretty much um blind
to what happens to the ultrasound wave
between these two moments and in order
to create a proper Imaging the whole
software the ultrasound wave needs to to
to make some assumptions in order to
process and create a proper um image
from the from the from the structure so
normally the ultrasound wave has four
creates four different assumptions first
of all the ultrasound wave propagates in
a straight line the second one that the
structure is going to be image just once
uh um the third one is the just that the
structures in the path of the beam are
going to be are going to be image and
are going to be generating the the
reflection and also that the position of
the structure is prop uh is proportional
to the travel time every time one of
these um assumptions are violated uh we
have the creation of one
artifact probably the most common
artifact that we see on on our daily
practice is the reverberation artifact
the way I like to think is that like the
ultrasound wave gets like trapped back
and forth between two different strong
Reflections as we can see on this
example here we have the ultrasound wave
from the transducer it's emitted and
then hit the a strong reflector bounces
back to the transducer that in most of
time can be can act canect as um a
strong reflector bounces back to the
first reflection keep doing this pinging
pong thing um and because the ultrasound
when it thinks that distance is
proportional to the time of reflection
it pretty much creates the image the
image structure like a with double the
distance from the original reflection
reflector as you can see on this image
you can see probably the strong
reflector on this image is the interface
between the ascending aorta and the
pulmonary artery and you can see see
there is a there is a reverberation
artifact created which is double the
distance from the original reflector
this one is easy to identify we all know
that probably this is an artifact also
one thing U that help us to identify and
to be sure that this is an artifact is
that the reflected um artifact moves uh
which a much higher amplitude than the
original structure the this one is easy
to identify but this one which is the
same artifact not not so much this
patient came for came from home from uh
for an outside U for an outside Echo for
a followup from for another disease and
it was seen on the descending aorta um
this kind of structure that could be
related to a um type A dissection of
course the patient the clinical context
the patient had nothing was not was
symtomatic and as you can see one tiip
to identify that this is like an
artifact is that with color um the
reverberation doesn't change anything in
terms of the color flow there are other
ways of the reverberation that the names
are different for example we have the
com tail artifact normally we seen the
descending a it's caused by is um is AR
arcero PL Ma they're very close one to
each other and then you have this bounce
back and forth from the reflector uh
reflected wave that gives um origin to
this com tail artifacts another
reverberation artifact with a different
name is the ring down artifact that
sometimes we see when you have like a
residual bubbles especially on the left
atum and it's caused by trapping of the
small small spaces with fluid that are
surrounded by um residual air it causes
what is called um com um ring down
artifact we can also see the same type
of artifacts reverberation in 3G the
mechanism is is pretty much the
same and sometimes we deal with these
artifacts on our on our daily practice
as you can see here we have the B view
with a with a Venus canola and it's hard
to identify at first which one is the
true um canola and which one is the
artifact caused by reverberation of
course the reverberation artifact always
going to be in um um in terms of depth
is going to be a little bit deeper than
um than the original structure but we
can see that sometimes especially at the
beginning uh when you see these images
it's important to differentiate between
one or the
other the other type of um artifact that
you normally see is acoustic shadowing
this one is is much easier to understand
here we have the usual example of a
mechanical mitro valve but you can also
see acoustic shadow shadowing caused by
um atosc um calcium in the popular
muscles also in 3G the same way we have
this huge PL in the AR root and also the
same thing with with the the SAS ort
another type of artifact that we we
don't see very often is much more common
when you're doing trans thoracic Echo is
refraction is when you have the creation
of um of um um the creation of the same
of the same image twice as you can see
on this mitro valve lifet looks like we
have a duplication of the micro valve is
caused by this type of artifact and and
lastly in terms of wave reflection and
refraction we have the M mirror Imaging
artifact we also see most of the time we
Imaging the the same Taska one thing
that is important to mention is that
every time we have a mirror Imaging
artifact color flow and post wave dopper
and continuous wave dopper they are also
mirror Imaging like the like the
original
image another type of artifacts they are
created by uh ultrasound beam properties
the most common example is the side lobe
artifacts what we what usually happens
we have the main lobe the main being and
then we have a small beings with lot
less energy laterally to the main being
so sometimes these lateral um being with
less energy they heat a very strong
strong reflector and then this wave is
reflected back to the transducer one of
the assumptions is that everything is
reflected to the ultrasound being is
interpret interpreted by the ultrasound
by the ultrasound system is coming from
the main beam so the image that is
laterally um um disposed is now is image
at the center of the screen we can see
this example here of the desending tasa
ort we see the guide wire in the
desending tasa and you can see this
radial Imaging um in the far field which
is kind of the side lob of this image
another example in the ascending tasa
ort the mechanism is the same and this
one is a true dissection but you can see
the AR caused by some side lobe artifact
it's important to differentiate for a
true
dissection of course the presence of
cardiac devices and also what happens in
the is another source of um Imaging
artifacts as you can see this electroc
calter image here easy to make the
diagnosis and of course we have
artifacts in 3D the most common and the
most uh easiest one to identify is this
teaching artifacts when you're using
like a multi beit acquisition but we
also have other artifacts that sometimes
we don't um pay too much attention when
you see the suture line for example of
this mitro ring um much thicker than it
it really is it's called blurring it's
caused by the difference between
resolution in the three different planes
the Axel uh sorry the axal lateral
elevation resolution this difference um
makes the the stitching artifact the
stitches of the mitro Ring much thicker
than the really are and blooming is
exactly the same mechanism but happens
with mechanical metallic
structures we also have we also have um
another artifact in 3G where you can see
the ortic valve um in short axis view
looks like there is a hole at the center
of the of the Artic valve especially in
diine this a Dropout AR the fact that is
easily um recogn recognized using the 2G
image which offers you a much um much
better Axel and lateral
resolution most of the artifacts are
easy to identify most of the time they
have no clear attachment they're not
reproducible in different views which
that's why it's important to always look
um one image that you don't know what
you looking at in different views and
most of and they are not affected by
color do full wave Doppler or continu
wave Doppler how to avoid most of the
artifacts change the angle of incidence
change the view change the ultrasound
settings most of the time you'll be able
to to to avoid this traps Imaging pitch
Falls they include not only artifacts
but they also include misinterpretation
of a properly represented structure
weather Normal or pathologic you you
know I'm pretty sure about all the speit
Falls we we have the station valve we
have the Kary Network which is sometimes
easy to um
to misrepresent as um vegetation on the
right side we also have um um Crystal
terminales we have lipomatous
hypertrophy of the interal septum all of
this um pit BS I'm sure we all know how
to identify we also have the moderator
band on the right ventricle
and we also have the kumag Gen reach
between the pulmonary vein and also the
left trage we have the tabulations also
of the left trage okay so for the last
part of my presentation I'm going to
present a few cases here they are
related to some artifacts there some of
them are related to some U to some pach
FS but it's important that the reason
why I brought these examples is because
first of all we found we we found them
on our clinical practice and it can also
happens with you and we can we need
always have to we need always to have a
high index of Suspicion in order to try
to avoid this traps this is the first
case this patient came in the morning he
had um tissue um a tissue mro valve in
the morning he was in the ICU um
overnight he became unstable he had a te
done and as you can see the LV of course
doesn't look good but if you look the
right ventricle looks like as a
collection on the right ventricle um
perhaps um pressing the right ventricle
uh impeding it um diastolic um its
diastolic feeling so it was made a
diagnosis of tonad or was called the
surgeon was called the patient was about
to go to the O but then reviewing the
morning images we can see the imag is
already there since the first image you
can see a normal B ventricular function
and you can see is H like to me a a very
big pericardial fat on the right
ventricle coming off bypass the
structure was also there even on 3G was
possible to see there was that it was a
picardal fat so or was canceled the
patient was treated um clinically and he
improved overnight and next morning case
number two as you can see patient came
with an acute Artic syndrome patient had
has um a pretty big intramural hematoma
as you can see also you can see on the
Des toas very close to the arch U there
is a dissection as we all can see but
what was really interesting at the start
of the case that it was difficult to
acquire proper images of the left
ventricle and even from the right
ventricle there are something um coming
in and out of our field it was really
difficult to see anything else looking
the turning the probe to the left
what we saw at the left uh um chest was
this um huge amount of fluid um we told
the surgeon but then look closely we
notice we cannot see the desending tasa
orta so we made a diagnosis of gastric
fluid we put an mg we drain almost 1.5
liters of fluid and then the image got
much better much better we were able to
finish our
examination this is a patient that came
for a redo um is anomy he had a tissue
tissue Artic valve done a few years um
back and he came for um severe Artic
insufficiency the valve need need need
to be replaced but the on the pre
nothing was mentioned about Mr we saw
this image the surgeon saw this image we
kind of a star in the case and he asked
us to quantify this how bad was this Mr
but if you look closely and I'm going to
show you the image without color you can
see the Mr it happens when the mitro
valve is still open so it was made the
diagnos of diastolic M which is kind of
described um happening in cases of
severe AR regation this the U frame that
I took um both images and you can see
the Mr happens and the micro valve is to
open so nothing was made with the mitro
valve patient went on bass had a tissue
valve um Artic valve replace and he did
well so left appendage always uh
important to make the diagnosis of um
low flow causing um smoke or even
causing like um creating some trombos
this is easy to make the diagnosis but
what happens if you have a patient
coming for a cardiov version you do a to
rule out um left atal left atend trobos
troms and you see this image would you
cancel the case send patient home for a
few more weeks of uh anticoagulants to
bring him back for another Echo and then
shock it's better if you stop you do a
comprehensive examination in this case
we saw that this image on the left was
actually um Pino muscles from the left
at appendage so we had we we finished
that EO patient had the cardiov and went
home on the same
day left ATR appendage thrombos this is
most of the time straightforward to make
the diagnosis you can see the left page
thrombus where it's supposed to be this
the same patient why on bass we had the
dislodged thrombos we told the we told
the surgeon he open the left at and
remove the left at um the trombos but
what we do when you have a case like
this patient came to come to the water
for another procedure um bypass almost
finishing we are kind of starting to
come off bipass and you see this image
patient had no factors for thrombos
looks like it's attached to the left
atal wall what do you make of of this so
would go back would you go back on
bypass and open the left at to see what
it is so this is very unusual to see but
it's possible to happen this an
invagination of the left appendage we
came off bypass the heart was full again
and the the left appendage went back to
his original place my last case this
patient had a trit uh replace
um a few years back due to endocarditis
he left the hospital and um he lost
followup for 18 months he came back to
our service this time um he had a
transic echo done somewhere else and it
was seen he had severe tros ped stenosis
he came for a reduced anotomy and tros
ped valve um replacement this is our
first image you can see there is
something on the crit valve without
colorist you see with colorist you see
you can see the flow acceleration
looking closely and with more detail
looks like there is something on top of
the trit valve kind of hard to see what
it is so this was done like quite a few
years ago so we did a like um 3D Zoom of
the trus pit valve and you can see of
course there is some stitching artifacts
but there is something on top of the the
TripIt valve kind of preventing it from
opening closing properly we went on
bypass and the reason of this patient
trus regation is that the holder of the
trus valve was still there patient had a
trit valve um replacement and he did
find uh coming off bypass so my
take-home mches are every time you see
something that is suspicious that you
think that is something that was not
supposed to be there first of all stop
and think course relate with the
clinical finds correlate with the
patient patient um clinical history if
you're still not like 100% sure or it's
happening try to acquire and try to
interpret the image using multiple views
in different modalities this is
essential for the for the final
information while doing the the te and I
try to use those cases just to examplify
situations in in which one error could
have led to different surgical results
and with that I'll finish my
presentation thank you again for
watching thank you Amari