0:02

Hello and welcome to Noon Conference, hosted by modality

0:05

Noon Conference connects the global radiology community

0:08

through free live educational webinars that are accessible

0:11

for all and is an opportunity

0:13

to learn alongside top radiologists from around the world.

0:16

Today we are honored to welcome Dr.

0:18

Dane for a lecture entitled Photon Counting,

0:20

CT Physics and Applications.

0:23

Dr. Dane completed her radiology residency

0:25

and body MRI fellowship at NYU Langone Health

0:29

where she also served as chief resident.

0:31

She's now an abdominal radiologist at NYU

0:34

where she also serves as director of CT

0:36

and the director of Quality and Safety

0:38

for Maine Campus Outpatient Imaging.

0:41

At the end of the lecture, please join her in a Q

0:43

and A session where she will address questions you may

0:46

have on today's topic.

0:47

Please remember to use that q

0:48

and a feature to submit your questions so we can get to

0:51

as many as we can before our time is up.

0:54

With that, we're ready to begin today's lecture. Dr.

0:56

Dane, please take it from here.

0:58

Alright, thank you for having me.

1:01

So I'm excited to talk to you about photon counting ct.

1:05

So this is the general breakdown for

1:08

what we'll talk about today.

1:09

We'll start by talking about the detector basics.

1:11

So that really covers the physics

1:12

and really what's important,

1:14

what plays out into the clinical applications.

1:16

So I am a clinical abdominal radiologist, so a lot of

1:19

what we talk about will be clinical work today

1:21

and then at the end we'll conclude

1:22

with some PACS considerations.

1:24

So first, the detector basics.

1:28

So here's a schematic

1:29

of a conventional energy integrating CT detector.

1:32

So most of the scanners on the market

1:34

that are not photon counting ct, you can see

1:36

that it has a scintillator.

1:38

So when the x-ray photon comes in, it gets converted

1:40

to light in that scintillator,

1:42

then the light gets converted to current.

1:44

So it's a two step conversion using light,

1:47

there are a number of consequences of using light

1:49

and one is that like a flashlight light can travel

1:51

in any direction.

1:52

So you need these opaque SEPTA in order to avoid crosstalk

1:56

between the adjacent detector elements, which in doing so,

1:59

does limit the spatial resolution

2:00

and dose efficiency of conventional ct.

2:03

And here's why. You can see with decreasing pixel size,

2:06

you have more and more opaque SEPTA

2:08

and more and more dead space.

2:10

Eventually it becomes too dose inefficient

2:13

to get smaller pixels

2:14

and that limit for most

2:15

of the conventional scanners in the market is about

2:17

0.6 millimeters.

2:20

So there are other consequences of using light

2:23

and one is that light is really fast.

2:25

So even the fastest detectors in the market cannot separate

2:28

the signal from different light pulses.

2:30

So instead it'll create a cumulative measure of all

2:33

of the energy in that detector.

2:36

But in essence that down weights, low energy photons,

2:39

unfortunately the low energy photons are the ones we wanna

2:42

see more of Those are closer to the K edge of iodine,

2:44

which means that iodine looks brighter on those.

2:47

The low energy photons are the ones that contribute most

2:50

to iodine image contrast.

2:52

So to give you an analogy, if I think

2:54

of conventional detectors, I think of them

2:56

as a bag of change.

2:58

It has some combination of coins, pennies,

3:00

nickels, dimes or quarters.

3:01

We have no idea what combination is in the bag for ct.

3:04

We hope there's a lot of low energy photons

3:06

because it increases iodine contrast to noise.

3:08

But we don't know. But we know this bag has 77 cents.

3:14

Now let's move on and we'll talk about this is the

3:16

clinically available photon counting detector.

3:19

So now you see it has a semiconductor

3:21

instead of that scintillator.

3:23

So when the x-ray photon comes in,

3:25

it now gets directly converted

3:27

to current in a single step conversion.

3:30

So without the scintillator there's no light.

3:32

Without the light, you don't need the septa.

3:34

Without the septa there's better dose

3:36

efficiency and spatial resolution.

3:38

So that we talked about less dead space.

3:40

So in addition, something else exciting about this is

3:42

that each x-ray will produce a

3:44

charge proportional to its energy.

3:46

So what does that mean? It means

3:47

that we can measure the energy

3:49

of every photon of the detector.

3:51

So let's go back to our analogy for conventional detectors.

3:53

We can think they're like a bag of change.

3:55

There's some combination of change in the bag.

3:58

You don't know if you have quarters or pennies,

4:00

but you know that there's 77 cents in the bag.

4:02

Whereas photon counting CT can tell you you have seven

4:04

pennies, five nickels, two dimes and a quarter.

4:07

So every coin is counted, every photon is counted.

4:10

So in this example we can see that we have seven coins

4:13

or that are pennies which account for 47% of the coins.

4:16

Whereas if you were to look at their energy

4:18

or their weight, 7 cents, seven out

4:20

of 77 would be 9% of the total.

4:23

So the same photons going into the photon counting detector,

4:27

they have greater weight to those low energy photons.

4:30

So there's greater iodine contrast to noise.

4:34

In addition, every scan is a multier energy scan,

4:36

it's detector based spectral imaging.

4:38

Finally, you can electronically remove the photons whose

4:41

energy is too low to have possibly penetrated

4:44

the body as noise.

4:46

Here's a schematic of another uh,

4:48

photon gowning counting detector.

4:50

This one uses silicon, which due to its uh,

4:54

low atomic number, it has an edge on geometry

4:56

and it has to be between three and six centimeters thick.

4:58

In order to achieve a stopping power in the diagnostic x-ray

5:01

range, it has eight energy bins, three of which are

5:05

below 33 KEV to capture scatter,

5:07

and then five of increasing energy

5:08

and height in order to ensure a relatively consistent

5:11

energy distribution.

5:14

So something that I think is really exciting about photon

5:17

counting CT is the CT number accuracy and reliability.

5:21

So I'll share with you two studies from the Wisconsin

5:23

group where they were looking at.

5:24

The first one was looking at patient size

5:27

and with varying patient size, they found

5:29

that the CT number was closer

5:31

to ideal at foton counting than dual energy

5:34

and single energy at various sizes with less variability.

5:38

And looking at size and positioning, well first they found

5:41

that size contributes more than positioning

5:43

to CT number fluctuations

5:45

and they found again, smaller changes in hounds field units

5:49

with greater accuracy with photon counting than

5:51

with dual energy and conventional ct.

5:54

So in general you're seeing increased CT number reliability.

5:59

So here's a table from a nice review article

6:01

that was published earlier this year in radiology

6:03

where they were uh, this, this table summarizes the

6:07

photon county detector scanners that are in development

6:10

with all the vendors and you could see

6:11

that many vendors are working on it.

6:13

We're not gonna go through this whole table,

6:14

but I do wanna just highlight a few things

6:16

in this general area.

6:18

So first you can see most of the vendors are using some sort

6:21

of cadmium telluride combination.

6:22

There's one vendor that's working with silicon.

6:24

Something else. If you look at the implants,

6:27

spatial resolution, nearly all

6:29

of them are less than 0.2 millimeters.

6:31

That's because of the absence of the opaque septa.

6:34

And the last thing I wanted

6:35

to highlight is the number of energy bins.

6:37

So conventional dual energy CT dual is two, right?

6:41

But here most

6:42

of these vendors have more than two available VIN bins,

6:45

which brings up the potential

6:47

for multier energy imaging in the future.

6:50

Okay, so now let's talk about the clinical applications.

6:53

We'll discuss them in order of these clinical, UH,

6:55

or physics related scanner related,

6:57

uh, benefits that we talked about.

6:58

So there's higher spatial resolution

7:00

because you don't have those opaque septa.

7:03

There's higher iodine contrast to noise

7:05

because there's greater weight given to low energy photons.

7:07

Also noise removal at the detector

7:09

we just discussed the noise reduction

7:11

and then spectral imaging is always available.

7:14

So let's start with spatial resolution.

7:16

So we'll do this more or less head to toe.

7:18

So we'll start in the temporal bones, which is a huge area

7:21

of opportunity for this improved spatial resolution.

7:24

Here's two images in the temporal bone in the same patient.

7:27

The left is photon counting. The right is conventional ct.

7:30

This patient has a cholesteatoma, which I've annotated.

7:32

But looking at these images, you can see there's really

7:34

better image quality

7:36

and less visible noise on the photon counting than the

7:38

conventional ct.

7:40

This study reported superior spatial resolution

7:43

and better critical structure visibility

7:44

with significant radiation exposure reduction at photon

7:47

counting compared with conventional ct.

7:51

Another temporal bone example,

7:53

these are two images in the same patient.

7:54

The left is photon counting the right's conventional ct.

7:57

You can see this cochlear implant

7:58

to better advantage on the photon counting than you can on

8:01

the conventional CT with less artifact.

8:03

And as expected, it's been reported

8:04

that photon counting CT does have a more precise

8:07

postoperative cochlear implant.

8:09

Uh, electro contact determination

8:11

with the reported substantial advantages compared

8:13

with conventional ct.

8:17

So a real practice changing area

8:20

of the spatial resolution

8:22

and iodine contrast to noise at photon counting CT are the

8:25

detection of CSF venous fistulas.

8:27

So here's a study where patients had an MRI with some degree

8:30

of probability of CSF venous fistula

8:32

but it wasn't detected on the MRI

8:34

and then they underwent a photon counting CT myelogram.

8:37

And in those patients you can see there's a drastic number

8:39

of these patients that had definitive CSF venous fistulas

8:42

identified on the photon counting CT myelogram

8:44

and it increased with degree of suspicion on MRI.

8:47

So this really is a practice changing area, excuse me,

8:51

of uh, photon counting ct.

8:53

There's some centers that are really uh,

8:55

really doing a tremendous job with this for their patients.

8:59

So now moving down to the lungs,

9:00

interstitial lung disease is another disease process

9:03

that really benefits from the spatial resolution.

9:06

So here we have two high resolution images

9:08

from the lungs in the same patient.

9:09

The left is photon counting CT high pitch pitch

9:12

of 3.2 millimeter slice thickness,

9:14

whereas the image on the right it's 0.6 millimeters.

9:17

So the highest we can get on that conventional CT scanner,

9:19

which it happens to be a very excellent scanner.

9:22

If you look at these images, you can see improved visibility

9:25

of the architecture of this honeycombing, uh,

9:27

or these cysts at least uh,

9:28

on the photon counting than you can in the conventional ct.

9:30

You can see the architecture more clearly

9:33

and as expected, it's been reported in the literature

9:35

that ultra high resolution photon counting CT does have a

9:38

more precise depiction of interstitial lung disease.

9:40

CT features with significant radiation exposure reduction

9:44

compared with conventional CT In terms of detect detection

9:48

of nodules and airways.

9:50

Uh, this group did a study looking at the barely detectable

9:54

nodule and barely detectable airway for both photon counting

9:57

and conventional ct.

9:59

And as expected you can see statistically significant uh,

10:02

smaller nodules

10:03

and airways are detected at photon counting ct.

10:05

And I realized this isn't a micro nodule,

10:07

I just thought it would be a little bit dry

10:08

to put a micro nodule on a slide.

10:12

So moving on to the abdomen, these are two axial images

10:15

through the kidney and the same patient both on

10:17

photon counting ct.

10:19

These are from the same exam and localized to the same slice

10:22

and the image on your left is four millimeter slice thick

10:24

and you can see that there's a smudge

10:26

of a non obstructing stone which would be very easy

10:28

to overlook and you really wouldn't be so confident

10:31

that's there compared with other areas of image noise.

10:33

But if you look at the very thinner,

10:35

the very thin slice 0.2 millimeter,

10:37

you can clearly see the stone.

10:38

You're confident it's there. Uh, really no debate here.

10:42

So as expected, it's been reported that there's greater odds

10:44

of detecting stones at photon counting compared

10:46

with thin section conventional ct

10:48

and there's significantly higher sensitivity

10:51

for renal stones at the thin section 0.4 millimeter compared

10:54

with one millimeter, especially the

10:55

detection of small stones.

10:57

So really important for people with stone disease,

10:59

we can give a more accurate

11:00

assessment of their stone burden.

11:03

Now in terms of pancreatic cyst,

11:05

there's also improved visibility of pancreatic cysts,

11:07

which I think is due to a combination

11:09

of the improved implant spatial resolution

11:11

and also the iodine contrast to noise.

11:13

So here's two contrast enhanced images

11:15

through the abdomen in the same patient.

11:16

The left is photon counting. The right is conventional ct.

11:20

Looking at the photon counting ct,

11:21

you see there's a six millimeter pancreatic tail cyst.

11:24

You see it, you know it's cyst.

11:26

It's frankly not that interesting

11:27

and maybe you wish you didn't see it,

11:28

but if you look at the conventional CT on your right,

11:31

this is the optimally selected image from that scan

11:33

and you really frankly can't see it as well.

11:36

So this ability to see these cysts more clearly on on photon

11:39

counting has been reported in the literature

11:40

that there's greater pancreatic cyst visibility,

11:42

particularly for CT angiography which also leverages

11:45

that iodine contrast to noise improvement in terms of bones.

11:50

Here's two axial images through the ankle.

11:52

These are in the same patient at the same window with level

11:55

and kernel, but you can see the improved trabecular

11:58

visibility on photon counting and conventional ct.

12:00

And this is supported in the literature.

12:04

So if I were to think of an area where spatial

12:06

and temporal resolution are really married,

12:08

it's coronary CT angiography

12:10

and as expected photon counting ct coronary CT angiography

12:14

has been reported to have improved objective

12:16

and subjective image quality compared with conventional ct.

12:19

But we need better than just excellent image quality.

12:21

We need diagnostic performance.

12:23

So a study of one type

12:24

of photon counting CT reported 100% sensitivity

12:27

and 90% specificity compared with conventional ct.

12:31

So it was better, it was 75 and 50%

12:33

and this was using invasive coronary angiography

12:35

as the reference standard in a very high risk population.

12:38

Another study looking at a high risk population of patients

12:41

with severe coronary calcium and prior stents.

12:44

So really patients that are very difficult

12:45

to do an excellent coronary CT angiography on.

12:49

They reported that ultra high resolution photon counting CT

12:52

had 95% accuracy per segment in this population.

12:57

There's also another study

12:58

that reported strong inter reader agreement at photon

13:00

counting where it was moderate at conventional ct.

13:02

So not only only only are we accurate

13:04

but we're also reliable in our interpretation.

13:09

Finally, another study looked at instant stenosis assessment

13:13

and compared ultra high resolution photon counting CT

13:16

with invasive coronary angiography

13:18

and found 100% sensitivity specificity and accuracy.

13:23

So this is huge. This is practice changing

13:24

because there's some centers

13:26

that are foregoing invasive coronary angiography

13:28

and instead undergoing ultra high resolution

13:30

photon counting ct.

13:32

Uh, so it's avoiding an invasive procedure.

13:35

So keeping all of this in mind, there's a group

13:37

that was looking at patients

13:38

who underwent ultra high resolution photon counting, uh,

13:41

coronary CT angiography patients who had stable chest pain,

13:44

they assumed 15,000 patients

13:46

and reported a 19% reduction in functional follow-up

13:49

and a 6% reduction in invasive coronary angiography.

13:53

And assuming a 10 year life expectancy per patient,

13:56

this resulted in SI $790 per patient saved,

13:59

which over the 10 year period was about $11 million.

14:03

So overall the ability

14:04

to reduce follow up saves uh, quite a bit of money.

14:09

Okay, so next higher iodine contrasted noise.

14:12

So this again was

14:13

because there's greater weight

14:14

to the low energy photons at the detector

14:17

and noise removal at the detector.

14:19

So whatever iodine is there you can see better.

14:21

So if you can see better, you can give less.

14:24

You also have improved visibility

14:25

of small vessels in small people

14:28

and there's improved lesion conspicuity particularly in

14:31

low contrast visibility areas.

14:32

So I'll show you examples of each of these now.

14:35

So here's two CT angiograms in the same patient.

14:38

The left is photon counting. The right is conventional ct.

14:41

You can see that we gave 70 ccs

14:43

of intravenous contrast on photon counting ct.

14:45

So 30 ccs less than the conventional CT

14:48

and the iodine conspicuity is quite similar

14:50

between the two studies.

14:53

Here's an example of a 50 cc uh, TAVR study.

14:56

It was a CTA chest, abdomen, pelvis and coronaries.

14:59

We only gave 50 ccs um, for CT angiography of the aorta.

15:03

There's a study that reported that a 25% intravenous

15:06

contrast reduction at photon counting CT had non-inferior

15:08

image quality compared with conventional ct.

15:12

So this phenomenon of reduced contrast has been pretty

15:14

well documented so far.

15:15

So for CT pulmonary angiography,

15:17

one study reported no reduction in image quality using only

15:20

35 ccs of intravenous contrast.

15:22

Another study, again,

15:23

no reduction in image quality at coronary CT angiography

15:26

using a 40% intravenous contrast reduction.

15:31

So all of these examples are CTAs.

15:33

So you're leveraging more dense contrast,

15:34

higher concentration contrast,

15:36

maybe leveraging low KEV reconstructions.

15:38

So what about venous imaging? Right?

15:40

This is the workhorse of what we do in abdominal imaging.

15:43

So here we have two axial portal venous

15:45

phase images in the same patient.

15:46

The left is photon counting ct,

15:47

the right is a conventional ct.

15:49

In the conventional CT we gave 95 ccs

15:52

of intravenous contrast, which is based on weight-based

15:54

dosing of 1.5 milliliters, uh, kilograms per milliliter.

15:57

But if milliliters per kilogram, sorry,

15:59

but if you look at the photon counting ct,

16:02

I'm showing you a 70 KV image.

16:04

So this has a similar appearance to

16:05

that conventional one 20 KVP scan.

16:08

So you're not leveraging low KEV reconstructions here.

16:11

Here we subtracted nearly 30 ccs off of that

16:14

and the iodine conspicuity is quite similar.

16:17

One other thing to highlight is the radiation exposure is

16:19

half and the image quality is still adequate.

16:22

So we did a study where we looked at these 70 KEV

16:25

reconstructions at photon counting CT

16:27

and showed that with a 20 uh cc reduction off

16:30

of weight-based dosing photon counting CT still had similar

16:33

attenuation and image quality as conventional ct.

16:36

Another group reported 27% intravenous contrast reduction

16:39

had maintained image quality compared

16:41

with second generation dual source dual energy ct.

16:44

And a third group looked at thorac abdominal CT

16:46

and again reported a 17% contrast reduction had superior

16:49

image quality than conventional ct.

16:53

So now small people and small vessels.

16:55

Uh, here's a patient who had a nine cc angiogram.

16:59

So really putting that into perspective,

17:00

it's about a test bolus

17:01

and you can, you can see it's an excellent study.

17:04

There was a study that was done

17:05

of 113 children all younger than three years old,

17:08

median 66 days

17:09

and they reported better image quality at the same

17:11

dose, which is quite low.

17:12

And they directly measured signal to noise

17:14

and contrast to noise and said it was statistically higher

17:16

than a dual source dual energy ct.

17:19

Here's a nice clinical example of

17:21

where this iodine contrast to noise is important.

17:24

So here's a six month old who had a prenatal lesion that was

17:26

concerning for sequestration.

17:28

This was a non sedated scan.

17:29

So really leveraging that high pitch imaging.

17:32

You can see on this sagittal image this is the celiac

17:34

artery, this is the left gastric artery

17:36

and you can see it's supplying this extra low bar upper

17:38

abdominal broncho pulmonary sequestration.

17:40

Bonus is its normal coronary origins,

17:43

again this is non sedated patient only received nine ccs

17:46

of intravenous contrast and the CTDI is 1.2.

17:49

So low radiation, very little uh, contrasting.

17:53

We can make an excellent diagnosis really due to

17:55

that improved iodine contrast to noise.

17:59

Another area that benefits from this is lower

18:01

extremity CT angiography.

18:03

So here's two axial images from below the knee

18:05

and these are the same patient one month apart.

18:09

You can see on the photon counting CT we gave 30 ccs less

18:12

contrast but there's improved visibility

18:14

of these small structures

18:15

and again, bonus lower radiation exposure.

18:18

There was a nice study outta the Mayo group

18:19

where they did same day photon counting CTA

18:22

and conventional CTA

18:23

and even with greater than 50% intravenous contrast

18:26

reduction at photon counting ct.

18:28

They still reported

18:29

that more fibular perforator arteries were identified

18:31

with improved arterial sharpness at photon counting compared

18:34

with conventional ct.

18:37

In terms of diagnostic performance,

18:38

this group was looking at patients

18:40

with peripheral arterial disease

18:41

and they reported 91% sensitivity, 95% specificity

18:45

and 93% accuracy for the detection

18:48

and diagnosis of peripheral arterial disease compared

18:51

with digital subtraction an geography.

18:52

So again, we're particularly accurate in our diagnoses.

18:58

So now moving on to the last category in this section

19:01

lesion conspicuity.

19:02

So I think this is most evident in the low contrast

19:05

visibility areas such as the liver

19:07

and the pancreas, which is what I'll show you now.

19:09

So here we have two axial portal venous phase

19:11

images in the same patient.

19:12

The left is photon counting ct,

19:14

the right is conventional ct.

19:16

I'm showing you a 70 K EV image at photon counting ct.

19:18

So not leveraging low K EV reconstructions

19:21

but I think we'd all agree we could see this metastasis more

19:23

clearly at photon counting than we can a conventional ct.

19:27

So it's not quite fair. These are abdomen window,

19:29

these are the same uh, window width and level.

19:32

So I'll put this into liver window

19:34

and now you can still see on this locate UV

19:36

image not even on liver window.

19:38

You can see still improved visibility

19:40

of this metastasis at photon counting.

19:41

Then conventional CT as expected, it's been reported

19:46

that at photon counting CT there's improved visibility

19:48

of these hypovascular liver metastases

19:50

and improved contrasted noise compared with conventional ct.

19:53

This is for low K EV reconstructions but also 70 DK ev.

19:57

So you don't have to use the locate EV

19:59

reconstructions to get that benefit.

20:01

So something that I think is challenging

20:04

with cancer studies is if you're comparing an old

20:07

to a new study, you never know if anything is new.

20:10

But something that I love about this study which is from the

20:12

Duke group which was published in radiology is they used

20:15

anthropomorphic phantoms and had 183 generated lesions

20:19

and then they scanned it with both photon counting

20:20

and conventional ct.

20:22

So now these are exactly the same lesions in both

20:25

of the scans they looked at the 70 KEV reconstruction

20:28

so you don't have to leverage low KEV

20:30

and they reported higher sensitivity with photon counting CT

20:34

and at low doses they had increased

20:36

sensitivity for small lesions.

20:37

So you really do see lesions better at photon counting ct,

20:40

which I think is due to that improved iodine contrast

20:42

to noise and also the in plain spatial resolution.

20:47

So the other low contrast visibility area

20:49

that I think is valuable is the pancreas.

20:52

So here's two axial pancreatic parenchymal phase

20:54

images in the same patient.

20:55

The left is photon counting. The right is conventional ct.

20:58

Looking at the conventional ct,

21:00

you can see there's an ill-defined uncinate mass

21:02

and I think the borders of it are difficult to discern.

21:05

If you look at the photon counting ct, again 70 KEV,

21:08

we're not leveraging low KEV reconstructions.

21:10

You can see the tumor more clearly than you

21:12

count in conventional ct.

21:14

We do have the ability to use low KEV reconstructions.

21:17

I did tell you at the very beginning

21:18

that spectral imaging is always available.

21:20

So here's a low KEV

21:21

where you can see a drastic improvement in the borders

21:23

of this tumor at photon counting compared

21:25

with conventional ct.

21:28

As expected there's improved pancreatic ductal adeno

21:30

carcinoma tumor conspicuity at photon counting compared

21:32

with conventional CT for low KV reconstructions

21:35

but again also at 70 KV.

21:39

So with the ability to see these tumors more clearly,

21:41

you'd imagine it would make us more reliable.

21:43

In our assessment we did a study

21:46

of four fellowship trained abdominal radiologists

21:48

and they had substantial interrater agreement

21:51

for celiac artery S-M-A-S-M-V

21:53

and common hepatic artery involvement.

21:55

Whereas three of these had moderate involvement at

21:57

conventional ct.

21:59

In terms of metastasis identification,

22:00

again we had substantial interrater agreement at photon

22:03

counting where it was moderate at conventional CT

22:05

and this is with about a 40% radiation exposure reduction.

22:11

Okay, so now let's move on to noise reduction.

22:14

So really noise reduction at the detector translates

22:17

to better image quality and lower radiation exposure.

22:20

So here's two axial portal venous

22:22

phase images in the same patient.

22:23

The left is spoke on counting. The right is conventional ct.

22:26

I think the image quality looks the same on these two

22:28

studies but you'll note that this is

22:29

with 50% radiation exposure reduction on photon counting ct.

22:34

This is well documented

22:35

for pretty much every exam type to date in the literature.

22:37

So I just picked a couple of highlights

22:39

in an oncologic cohort.

22:40

One study reported that a 32% size specific dose estimate

22:44

reduction at photon counting again had the same image

22:46

quality and noise as dual source dual energy ct.

22:50

Another group looking at TAVR studies which are pretty high

22:52

radiation in general reported statistically significant

22:55

reductions in radiation with better image quality at photon

22:57

counting compared with conventional CT

23:00

for CT pulmonary angiography.

23:02

This group reported nearly 50% reductions in radiation at

23:05

photon counting compared with conventional ct.

23:08

But a nice perk that they noted was

23:10

that spectral imaging was available in all patients in their

23:13

study at photon counting, whereas a conventional ct,

23:16

this was dual source, dual energy ct,

23:17

it was only available in 66% of the patients

23:20

because of the need for high pitch imaging in others.

23:24

Back to pediatrics.

23:25

So here was a 16 month old who had concern

23:28

for childhood interstitial lung disease.

23:29

Again another non sedated scan.

23:31

So you're leveraging high pitch imaging.

23:33

Here we only gave five ccs

23:34

of intravenous contrast CTDI less than one.

23:36

We see the lungs are normal.

23:38

So in terms of radiation, uh Marilyn Siegel's group,

23:42

they found statistically significant reductions in radiation

23:45

at photon counting compared with conventional ct

23:47

but no difference in image quality signal to noise,

23:49

lung attenuation or noise.

23:51

Clearly just documentation

23:53

of this noise reduction which benefits

23:55

our pediatric patients.

23:58

So now moving on. So uh, another group was looking at

24:01

low dose kidney stones.

24:03

So most institutions have some sort

24:04

of low dose kidney stone protocol

24:06

and even off of already low dose they showed 44%

24:10

radiation exposure reduction.

24:12

They've marry, uh, measured the signal

24:14

to noise ratio at various portions along the collecting

24:17

system such as in the kidneys at the level of the

24:19

so acid at the level of the operators

24:20

and showed higher signal

24:21

to noise at photon counting than conventional CT

24:24

and no difference in radiologist confidence.

24:26

So really with the noise reduction we can reduce radiation,

24:29

we get better image quality, better signal to noise,

24:31

better contrast to noise

24:32

and no change in our confidence for diagnoses.

24:36

So a group of patients that particularly

24:38

benefits from this noise reduction or obese patients.

24:41

Here we have two axial CT angiograms in the same patient.

24:44

The left is photon counting.

24:45

The right is conventional ct other than decreased size

24:48

of this pancreatic tail collection on the subsequent photon

24:52

counting ct, I think the image quality is quite similar

24:54

between these but again this is

24:55

with a 40% radiation exposure reduction on photon counting

24:58

CT in obese patients,

25:01

one study reported 25% CTDI reduction again had similar

25:04

to improved image quality at photon counting compared

25:07

with dual energy ct.

25:10

So something else that I like about photon counting CT is

25:14

that spectral information is available

25:15

for the entire scan field of view.

25:17

So here's two completely uncropped images

25:19

and on photon counting CT you can see this iodine

25:22

information available everywhere whereas in the same patient

25:24

the patient is slightly off ISO center,

25:26

you only have the spectral information in the

25:29

smaller B tiled field of use.

25:30

So you're missing quite a bit of this patient's anatomy.

25:34

So now to show you a clinical application

25:36

of this noise reduction, here's three thin images uh,

25:39

in the same patient all on photon counting ct.

25:42

The left two images are the same exam in the

25:44

right as a prior study.

25:45

So with the noise reduction we really can use sharper

25:48

kernels than we've been able to

25:50

for clinical purposes At conventional CT

25:54

here these are localized to the same exam.

25:55

You can see this right lumbar artery

25:57

and actually you can see it's supplying this tiny type two

26:00

endoleak and it's literally blurred out

26:02

by the blooming on the softer kernel from the same exam.

26:06

These sharp kernels have been shown

26:07

to improve stent lumen visibility and sharpness

26:10

and compared with digital subtraction angiography,

26:12

radiologists have had greater confidence.

26:14

They reported improved vessel definition,

26:17

rep reduced calcium blooming and still acceptable noise.

26:23

So now let's talk about spectral imaging.

26:25

So spectral imaging is always available.

26:27

You have the gamut of what you're used to.

26:28

It's really detector based spectral imaging.

26:30

So here's a nice clinical example of how you can use it.

26:33

These are three images in a uh,

26:35

from the same exam localized to the same slice.

26:37

This patient had a colon cancer metastasis

26:39

in their liver ablated.

26:40

Looking at the 70 K UV image,

26:42

I can clearly see there's no recurrence here,

26:44

but I really love these iodine maps

26:46

because it's really like a black hole.

26:48

You can see there's no recurrence here,

26:50

there's no debate in the decision.

26:53

So in this section I'll talk about the virtual non-contrast

26:55

images and iodine.

26:56

We've already sort of talked about the

26:58

virtual monogenic images.

27:00

So the virtual non-contrast images do

27:02

have excellent image quality.

27:04

Here's two portal venous phase derived virtual non-contrast

27:07

images in the same patient.

27:08

The left is photon counting. The right is conventional ct.

27:11

I think we would agree that there's better image quality on

27:13

the photon counting CT with visibly less noise

27:15

and you could see improved uh, visibility

27:17

of these small structures such

27:19

as these gastro hepatic ligament vessels.

27:21

And this is with a quite substantial

27:22

radiation exposure reduction.

27:24

And we did document this in the literature

27:27

so we do need more than excellent image quality.

27:28

We need hounsfield unit reliability.

27:30

So here's a true non-contrast

27:32

and a virtual non-contrast in the same

27:34

patient and then I copied it.

27:35

They were quite similar frankly I copied

27:37

and pasted ROIs here so they'd be the same location and size

27:40

and you see that the household units are

27:41

effectively the same.

27:44

These are older studies

27:45

but one study reported

27:46

that the virtual non-contract household unit errors were

27:49

less than five household units in 76% of patients.

27:51

So really within calibration are

27:53

and less than 10 household units in 95% of patients.

27:56

In terms of diagnoses for hepatic steatosis,

27:58

there was an a UC of 0.97 for virtual compared

28:01

with true non-contrast.

28:02

So really it does quite well.

28:04

And if you were to use a liver

28:05

to spleen hansfield unit ratio less than 0.96,

28:08

it had 95% sensitivity

28:09

and nearly 100% specificity for the diagnosis.

28:14

So still, uh, speaking about steatosis here was a study

28:17

of over 500 patients, over 100 of whom had macd

28:20

and they compared the folks on counting fat fraction

28:23

with M-R-I-P-D-F-F.

28:25

Ultimately their conclusions were

28:27

that the CT fat fraction was both reliable

28:29

and precise compared with M-R-I-P-D-F-F

28:31

and that it was reliable across various radiation exposures

28:35

and tube voltages.

28:38

So we've had spec imaging for quite some time.

28:40

That's not really new

28:41

but something that is exciting is that we can leverage it

28:44

with high pitch imaging.

28:45

So here's images from a TAVR study.

28:48

So high pitch, you can see this is the contrast enhanced

28:50

image 70 KV

28:51

and you see an 83 hounds field unit left renal lesion

28:54

on a single phase study.

28:55

This would be in determinate, right?

28:57

We would say this is a hyperdense or a mass.

28:59

It would need some sort of follow up imaging.

29:01

I don't know if your institution recommends

29:02

ultrasound or MRI.

29:04

Mine usually recommends MRI but we can do better than that.

29:07

Here's the virtual non-contrast derived from the same image

29:10

where you can see it's 82 hence reel units.

29:12

So we know this is a hyperdense, the thought

29:14

of tumor never reaches the report,

29:16

neither does the recommendation for follow-up imaging.

29:19

The virtual non-contrast images when derived from both the

29:22

arterial and venous space have been shown to be accurate

29:24

and reliable for renal cyst characterization.

29:27

So another common scenario, this is a patient

29:29

who has lung cancer and adrenal nodule.

29:31

It's 31 hounds field units on this contrast enhanced image.

29:35

So as we know, lung cancer does have a propensity

29:37

to metastasize to the adrenals.

29:39

So on a single study, single phase study with no priors,

29:42

you're unfortunately left with could be an adenoma

29:45

or a metastasis and it would most definitely be recommended

29:47

for a follow-up study whether MRI

29:49

or pet, uh, whatever your institution does.

29:51

Mine usually does MRI.

29:54

But here's the virtual non-contrast from the same image

29:56

showing it to be 10 hansfield units.

29:58

So we know this is an adrenal adenoma, the thought

30:00

of metastasis never reaches the report, neither does

30:02

that recommendation for follow-up imaging.

30:07

Uh, here's a study that was looking at the difference in

30:09

virtual non-contrast attenuation

30:10

between adenomas and metastases.

30:12

And they showed statistically a significantly lower

30:15

attenuation in the adenomas and metastases

30:18

and they reported a 26 handhold unit threshold had 87%

30:21

sensitivity and 76% specificity for this differentiation.

30:27

So a big question is can the virtual non-contrast uh,

30:30

reconstruction replace a true non-contrast?

30:32

Well here's a single phase CTA

30:34

and a patient has an endograft and you can see

30:35

that there's a small hyperdense focus anterior

30:37

to the endograft which is not present on the arterial drive

30:41

virtual non-contrast.

30:42

So even though this is a single phase study,

30:43

I'm quite confident that this is an endoleak

30:46

and it has been reported

30:47

that there's comparable endoleak detection on image quality

30:49

comparing virtual versus true non-contrast images.

30:52

So something else that I wanna highlight is

30:54

that there's actually excellent iodine removal here on the

30:56

virtual non-contrast image even though it's from a dense

30:58

phase of contrast, which I think is something

31:00

that the dual energy CT systems have really

31:02

had difficulty with.

31:05

So keeping that in mind, something that we uh, experimented

31:07

with at RX institution was a graphic derived virtual

31:10

non-contrast image in our young patients to avoid radiation.

31:14

So here's an example of the urogram

31:15

and here's the graphic derived virtual non-contrast image

31:18

where you can see there's excellent iodine removal

31:20

and we could still see that bladder stone.

31:23

So we did study this before doing it of course,

31:25

and we found that at one 40 kv, which is

31:27

what we do these at the median collecting system

31:29

attenuation was 0.9.

31:31

So really close to zero.

31:32

And we had three fellowship trained abdominal radiologists.

31:35

82% of their scores were five outta five

31:37

for calculi detection.

31:40

So something else that's new is the ability

31:43

to leverage spectral imaging for coronary CT angiography.

31:46

So on your left is a true non-contrast

31:48

and the right is a virtual non-contract intended

31:51

for a calcium score from the coronary ct.

31:53

I think the calcium deposits look quite similar on these

31:56

images, but obviously we need more than that.

31:57

We need diagnostics.

31:58

So here's a study in radiology of 170 patients

32:01

where they reported no difference in a Augustine scores

32:03

and strong correlation between the true

32:05

and the virtual non-contrast.

32:06

So perhaps in the future maybe we'll see

32:09

that true non-contrast, uh, replaced

32:11

by the virtual non-contrast,

32:12

but I think uh, more studies will be performed.

32:15

Here's another example.

32:16

These are spectral images from coronary CTA.

32:18

This is the iodine map

32:19

and this is the virtual non-contrast image.

32:21

You can see there's no atrial linage clot

32:22

here quite confidently.

32:26

So moving on, this is something near and dear to my heart.

32:29

So iodine density. So why iodine density?

32:32

So iodine density, it's a parameter that's only attainable

32:35

for multier GCT

32:37

and it only reflects the iodine content within a voxel which

32:40

is from contrast.

32:41

Whereas hounds field units

32:42

or CT numbers, they vary based on intrinsic tissue

32:46

attenuation which will be incorporated

32:47

and they may vary with other exam related characteristics.

32:51

So we've been looking at iodine density as a marker

32:53

of con's disease activity.

32:54

So if you look at this image, you see the terminal ium,

32:56

it's thick walled, it's stratified, it looks bright, uh,

32:59

compared with the rest of the bowel.

33:00

So you already think this is active inflammation,

33:02

but if you like numbers we can draw do a mural region

33:05

of interest and you can see

33:06

that it's 3.6 milligram per milliliter.

33:09

What we did a study compared with endoscopic histopathology

33:11

where we found that 2.7 milligram per milliliter was a

33:14

threshold between active and no active inflammation

33:17

with 97% sensitivity and 100% specificity.

33:20

And we even found that we could differentiate

33:22

inflammation severity.

33:23

So an iodine density of 3.4 milligram per milliliter could

33:26

differentiate mild from moderate to severe inflammation,

33:28

again with excellent diagnostic performance.

33:31

So this patient's iodine density is 3.6, which is

33:34

above our 3.4 milligram per milliliter threshold.

33:36

So it correlates with moderate to severe active inflammation

33:39

and that's concordant with endoscopic

33:40

histopathology from two weeks prior.

33:44

Another group was looking at iodine concentration

33:46

as a marker of neoadjuvant therapy response

33:48

and here they drew free hand region

33:50

of interest measurements bound

33:51

around rectal tumors on the iodine map

33:53

and they had an A UCF 0.85.

33:55

So I think this is exciting for the possibility

33:57

of iodine concentration as a biomarker of disease processes.

34:02

So finally the last thing in spectral imaging

34:04

that I'm excited about for the future.

34:06

So this is an animal model

34:08

and the animal was injected with a, uh,

34:10

it was a dual injection with iodine

34:12

and then gadolinium at different time points,

34:15

but a single imaging acquisition.

34:17

They then imaged the animal once and generated iodine maps

34:20

and gadolinium maps.

34:22

The iodine was time to be for the hepatic artery

34:24

and you can nicely see the branches

34:25

of the hepatic artery here, whereas the gadolinium was time

34:28

to be for the portal veins

34:29

and you can nicely see these branches

34:30

of the portal of means.

34:32

So I think this is something that's exciting for the future,

34:34

the possibility of eliminating uh, imaging acquisitions

34:38

by using multiple contrast agent imaging.

34:41

So is it ready for prime time?

34:43

So here's a phantom study recently published in A JR

34:45

where they looked at uh, gadolinium based contrast

34:48

and they showed that there was appreciable hepatic

34:51

enhancement at 200 micromoles per kilogram,

34:53

which is much lower than previously seen

34:55

with conventional ct,

34:57

but no enhancement

34:58

of observed at the clinically approved 25 micromoles per

35:01

kilogram uh, concentration.

35:03

So not quite ready, uh, for real life yet.

35:05

So now let's talk about some PAX considerations.

35:09

So here's four images of different kvs,

35:11

the same window within level 4,400.

35:14

So I showed you a few examples

35:15

of why we like low KEV reconstructions.

35:17

They increase lesion conspicuity in low contrast visibility

35:21

areas, but they also are noisier

35:23

and everything gets brighter.

35:24

So if I were to look at this reconstruction at this window

35:27

within level, it really is isn't that appealing to my eye?

35:30

Whereas this reconstruction on your right is,

35:33

and you know, it brings up the concept

35:34

of default window width levels.

35:36

So you probably have fewer photon counting cts than

35:39

conventional cts in your practice.

35:41

And you know you have two choices.

35:43

Do you want to integrate your scanner

35:45

in with the rest of your scanners?

35:46

Use your default window within level level settings,

35:48

then maybe use something like 70 K EV

35:50

or go all in on low K EV reconstructions

35:53

and have optimized window width

35:55

and level settings for photon counting ct.

35:58

Most people in the country have selected this 70 KEV sort

36:01

of same window width and level approach,

36:03

but that's an institutional sort of decision.

36:06

One other concept when looking at low KEV reconstructions is

36:09

the concept of pseudo enhancement.

36:11

So everything gets brighter on low KEV reconstructions.

36:14

So here we have a 70 K ev.

36:16

So similar to that one 20 KVP scan.

36:18

We have a low KEV at 55 KEV, these are the same window width

36:21

and level and then a virtual non-contrast.

36:24

So looking at this renal lesion on the virtual non-contrast,

36:26

it's five hounds field units.

36:27

So I know it's a cyst.

36:28

I copied and pasted the ROI onto each of these images.

36:30

So they're the same place

36:31

and size on the 70 KEV, it's 18 household units.

36:35

So this 13 household unit change

36:36

I would never say is enhancement.

36:37

I would be perfectly happy saying this is a

36:39

cyst, simple cyst.

36:41

But if I look at this low KEV reconstruction,

36:43

you see there's a 19 household unit attenuation change.

36:46

So for people that are sticklers to

36:47

that 20 household unit attenuation change rule

36:49

for enhancement, this would make you bat twice for sure.

36:52

So there definitely would need to be a larger threshold

36:54

for low KEV reconstructions.

36:57

So here's the axial single shot fast pin echo

36:59

and the post con T one post contrast subtraction image

37:01

showing that this is clearly just cyst the 70 KEV

37:03

and the virtual non-contrast didn't lie.

37:06

So there was a study looking at this concept

37:08

of pseudo enhancement in renal cyst

37:10

and they actually showed no pseudo enhancement for CIS

37:13

between 70 and 190 KEV,

37:15

but increasing degrees of pseudo enhancement

37:17

with decreasing KEV.

37:19

So really something to keep in mind when routinely viewing

37:22

low KEV reconstructions.

37:25

Finally, something that's really important is the concept

37:27

of image overload and you really have

37:28

to be thoughtful in building your protocols.

37:31

So something exciting for me as a spectral enthusiast is

37:33

that most people are routinely viewing low KEV

37:36

reconstruction, such as, I mean, sorry, uh,

37:37

virtual mono energetic reconstruction such as stone here.

37:39

So this is a complete sub in sub out,

37:42

but um, you know, you also could add a virtual non-contrast

37:44

and an iodine map in my center.

37:46

P people really like them and if they're not there,

37:48

you know, that would, that would be a complaint.

37:50

So I'm thrilled with that. And we, we do

37:52

that on our conventional dual energy scanners too.

37:54

So that's really not a difference for us compared

37:55

with conventional dual energy ct.

37:58

But I mentioned in the very beginning that you can go down

38:00

to 0.2 millimeter slice thickness.

38:02

So if you're routinely reading off of three millimeters,

38:04

for example, and then you start sending 0.2 millimeters,

38:06

that's a huge change for your practice.

38:09

For us, we routinely send thin images to PAX for all

38:12

of our scans for a variety of reasons.

38:13

They're usually at 0.6 to 0.8 on all of our scanners.

38:16

So if I were to go from 0.6 to 0.4 for example,

38:18

that's really not that big a change,

38:20

but you have to keep in mind, you can go down to 0.2,

38:23

but if you want spectral reconstructions

38:25

with the clinically available model, you have

38:27

to use 0.4 millimeter.

38:29

So would you send 0.2 millimeter and 0.4 millimeter to pax?

38:32

Clearly the answer is no, you wouldn't.

38:34

You have to be thoughtful in your protocols.

38:36

Do you really need the 0.2?

38:38

If you do like temporal bone

38:39

or interstitial lung disease, then use it

38:41

in abdominal imaging.

38:42

We use 0.4 because, uh, you know,

38:44

i i I value spectral imaging.

38:47

So that's it for our talk on FO count counting ct.

38:49

We started by talking about the detector basics, really how

38:52

that difference in the detector, uh,

38:54

really brought out the clinical applications.

38:56

So we talked about the higher spatial resolution

38:58

because there's no opaque SEPTA particularly valuable

39:01

for looking at temporal bones and interstitial lung disease.

39:03

For example, the higher iodine contrast to noise is

39:07

because there's greater weight through the low energy

39:08

photons, also noise removal at the detector

39:10

so you can see your contrast better.

39:12

So that's helpful for reducing contrast,

39:14

seeing small vessels and lesion conspicuity.

39:18

The noise reduction at the detector translates

39:20

to improved image quality, uh,

39:22

with lower radiation exposure.

39:23

Also really helpful in obese patients.

39:25

And finally, spectral imaging is always available.

39:27

You have the gamut of what you're used to.

39:28

We talked a lot about virtual non-contrast

39:30

images and iodine.

39:32

And finally for pax, when you're building your protocol,

39:34

please be thoughtful about what you need.

39:36

There's lots of choices and you, uh,

39:38

don't wanna overwhelm your radiologists.

39:40

So that's it for photon counting ct,

39:42

thank you for your time and attention.

39:44

Thank you so much Dr. Dane. That was wonderful.

39:46

We are now going to open up the floor to questions.

39:50

At the moment there is nothing in that q

39:52

and A feature, so we'll pause for a couple seconds.

39:55

Sometimes it takes a bit for folks to ask.

39:59

If you've got questions, go ahead and place them in that q

40:02

and A feature in the zoom

40:05

and we'll try to get to as many as we can.

40:08

The first question is,

40:09

are there any applications for neuroimaging?

40:12

Yes, there's a number of applications for neuroimaging.

40:15

So we talked about uh, temporal bones, we talked about uh,

40:19

CSF venous fistulas.

40:20

Those are, you know, big impact applications.

40:23

Um, I think vascular imaging in general is a huge

40:25

application because

40:26

of the improved iodine contrast to noise.

40:28

Uh, so many groups are looking at that as well.

40:30

And then finally, um, if there's any vascular stents, um,

40:33

I think you can see those more crisp and sharp as well.

40:38

Um, okay, our next question,

40:45

uh, sorry it's scrolling.

40:48

Uh, is about ways to offload MRI volumes.

40:52

Do I have insights to specific use cases

40:53

where photon counting can offer similar diagnostic

40:55

information as MRI, such as liver fat quantification?

40:59

So, um, liver fat's a really good one

41:01

because there's data that shows that it's comparable.

41:03

So I think one of the things that's interesting is photon

41:05

counting CT is very new.

41:06

So you know, the clinical comparisons with conventional CT

41:09

and MRI are still coming out.

41:11

There's, you know, yeah there's specific disease entities

41:14

that have been reported in, you know, in terms of uh,

41:18

diagnostics compared with MRI, um, it's not

41:22

as widespread yet as we'd like.

41:23

Um, but you know, I, I think yeah,

41:25

liver fat is a really good one right now.

41:28

Also, I mean there's plenty of patients

41:29

that can't undergo MRI, um, like pancreatic cysts.

41:32

I think we can do a great job with CT as well.

41:36

Um, what are some pitfalls or weaknesses of photon counting?

41:40

I think um, I think when you get a new

41:45

uh, scanner, I think

41:46

whenever you get any new scanner, it's sort of like a plug

41:49

and play, not photon counting ct.

41:50

You can in general take your protocols,

41:52

you put them on the scanner, you know,

41:54

you look at the images and you know, you check them

41:56

and it's really not that much interaction,

41:58

but foton counting CT is completely different

42:00

than all of your other scanners.

42:01

So it really does require, um, a lot more work

42:04

to develop your protocols, uh,

42:06

because there's a lot more choices.

42:08

And I think some of those things are institution specific.

42:11

Like for example, I got

42:13

to the point I I I mentioned the

42:14

radiation exposure and image quality.

42:16

It's not like there's one size fits all.

42:18

There's, you know, some institutions may choose

42:21

to have better image quality, uh,

42:23

and the same radiation,

42:24

whereas others may choose some sort of combination.

42:27

Um, so I think the adjustment

42:28

to the new technology was definitely, um,

42:31

was definitely uh, something to adjust to.

42:35

How long does it take to reconstruct photon counting ct?

42:38

Um, so today it keeps up with our practice.

42:41

So we have um,

42:42

our photon counting CT is in a busy clinical uh, practice.

42:47

We scan it's open for about 10 hours.

42:49

We scan 55 patients in that time.

42:51

Many of those are chest, abdomen, and pelvis.

42:53

Um, and it keeps up with the images on pax so um, you know,

42:56

keeps up at the moment.

43:00

Um, how long have I been using Photon County ct?

43:02

We got ours in 2022.

43:04

We were uh, among the first in the country to get it.

43:07

What QIR setting

43:12

are you using in your institution?

43:15

Um, R docs like QIR one

43:18

but that limits how much we can optimize radiation.

43:20

Yeah, so we usually do um, we do QIR of three or two.

43:25

So in abdomen we're using QIR three

43:27

and chest three using QIR two.

43:29

So one of those

43:33

mm well it's scrolling.

43:37

I'm sorry, what abdomen kernel are you using?

43:40

We're using 60 KEV bbr 40.

43:43

We've tried increasing to 70 but it looks a bit uh, soft.

43:47

Um, we like br uh, 44 actually we recently did a uh,

43:52

consensus in abdominal, uh, A JR from the um, SAR

43:57

of photon counting ETC,

43:59

and uh, we actually came to consensus on that BBR 44.

44:01

For what it's worth, do I prefer certain exams on

44:05

the photon counting ct?

44:07

Um, well as an abdominal radiologist I wish all

44:09

of my abdominal imaging could be there,

44:11

but that's obviously not uh, how life works.

44:13

So, uh, in general, cardiac is really a winner.

44:16

So patients that have stents or calcium

44:19

or high BMI for cardiac would go there, TAVRs must go there.

44:23

Um, our pediatric patients, we really try to get them there.

44:26

Um, you know,

44:27

and then the other applications that we've shown, like if,

44:30

you know, our scanner is at a site

44:32

where we also have a single energy conventional CT

44:35

and you know, like if a temporal bone comes, we would try

44:37

to put it there and you

44:39

know, okay.

44:43

I think there are three photon

44:44

counting scanners available now.

44:45

What are the differences? Can all do spectral imaging?

44:47

So yeah, so all can do spectral imaging.

44:49

So uh, the top model, uh, you know,

44:53

has a bigger detector, it's dual source.

44:55

The second model it's uh, four centimeters smaller.

44:58

The detector is still dual source.

45:00

The third model is single source.

45:02

So, um, all can do spectral imaging.

45:04

I think the only difference would really be cardiac.

45:05

Um, 'cause with the dual source scanners you'll get

45:07

higher, uh, pitch imaging.

45:12

Uh, do I think the applications in liver fat quantification

45:15

steatosis evaluation are useful?

45:17

Given how common MRI

45:18

and ultrasound are for that purpose, are there instances

45:20

where the immediacy of CT is worth the radiation

45:22

dose compared to those modalities?

45:24

You know, so that's a good question.

45:26

Um, not everywhere does, um,

45:31

not everywhere does ultrasound for example, for,

45:34

you know, quantification.

45:35

Um, many places do mr but not every patient can tolerate mr.

45:40

Um, you know, MR takes a long time. Mr.

45:43

Elastography is not always available.

45:44

Some patients have, you know, pacemakers

45:46

or things that make it really difficult.

45:48

So I do think it's worthwhile to look at ct especially

45:51

because we can get such low doses radiation exposures.

45:56

Um,

45:59

I think you got 'em all.

46:00

Got 'em All the questions we finished.

46:04

Thank you so much for this lecture

46:05

and thank you so much for hanging out

46:06

and answering all those questions.

46:08

There's one more that just came in,

46:09

I don't know if it's okay.

46:12

Um, what was your method

46:13

for optimizing the window level on low energy?

46:16

That's a great question. So, um,

46:18

with our pacs there's an auto window, um, auto window,

46:22

auto window tool where you literally just draw a circle

46:26

around whatever you want and it'll optimize it for you.

46:28

So that's what I personally use in practice.

46:30

Um, there's another site that is using optimized window

46:34

with her levels for low KEV

46:35

and they just send it routinely from the scanner at

46:39

that low KEV.

46:40

Um, I think, you know, one of the challenges of that is,

46:43

so I'm an abdominal radiologist, the scan starts at,

46:46

you know, at the lung basis.

46:47

So the first thing I would do would change it

46:48

to lung windows so I could look at the lung and,

46:50

and then I'd have to go back to abdomen.

46:52

So, um, it may be helpful to have, uh, you know,

46:56

some presets a hot key for it if

46:58

that's what you choose to do.

47:00

Um, that's not the path that we've chosen.

47:01

Um, there is a center that's that, that has done that

47:03

and they actually published a paper this week, um,

47:06

in European Journal of Radiology

47:07

with suggested optimized defaults.

47:11

Could all that liver fat quantification be done on routine

47:14

liver surveillance ct?

47:15

Uh, yeah, theoretically it can.

47:17

Um, which pax my using visage.

47:24

All right. Shall we call it?

47:29

Thank you so much. Thank no for your

47:30

attention and for your questions.

47:31

Uh, I really appreciate the, you know,

47:33

the engagement and participation.

47:36

Yeah, for sure. And thank you so much for this lecture.

47:38

It's been a big request.

47:39

So appreciate you filling a huge

47:41

educational need for our audience.

47:43

Glad to, glad to be here. Thanks for having me. Awesome.

47:45

Yeah, and thank you so much for everyone else

47:47

for participating in today's conference

47:49

and all our previous new conferences.

47:51

Um, be sure to look out for that email later today

47:54

with a link to the replay so you can watch us again.

47:58

Please join us next week, Wednesday,

48:00

November 26th at 11:00 AM Eastern.

48:02

Dr. Csh McCury is going

48:04

to deliver a lecture entitled Anatomy

48:06

and Pathology of the Lateral and Posterior Skull Base.

48:09

You can register for that@modality.com

48:11

and follow us on social media

48:12

for updates on future NOOM conferences.

48:14

Thanks again for learning with us and have a great day.