Youtube daily report Jun 20 2017

My name is Noah Vesely.

I am a software engineer

at Freedom of the Press Foundation.

I work on SecureDrop,

it's a whistleblower submission platform.

We're currently deployed in

20-something news organizations

mostly in the U.S.,

but we have a number of organizations abroad

who are also adopting.

This system is designed to be used

by news organizations

who wish to provide a platform

where sources who want to protect their identity

because they're potentially exposing corrupt acts

by maybe powerful people---

or not-as-powerful people---

but still want to ensure that

this source of information does not get tracked back to them

are able to pass along this information

and hopefully serve some public interest in the process.

To SecureDrop, an adversary could be

anyone from a local police department,

or government agency,

to a corporation that someone works for,

to really any sort of actor.

And we realize that there are some

very powerful, state-level adversaries

who have very good hacking capabilities.

GPG in SecureDrop allows us

to keep information on the web server

in an encrypted form

so that the web server,

only for very brief periods of time,

has any access to any plain-text material.

We've decided on GPG instead of other alternatives,

which we've had to weigh carefully,

in large part due to the fact that

GPG is the most well-known

and most used tool for asymmetric encryption.

GPG is a free software implementation

that is widely considered to be a very good implementation.

And the uniqueness over this,

besides using some library that might have some fancier crypto,

maybe like libsodium or something,

is that there are no end user tools

for those sorts of libraries,

whereas for GPG we can publish

the public key of this offline air gap

in which submissions eventually get decrypted

and sources themselves are able to download that

and use it to provide the best encryption---

like what we would call end-to-end encryption---

possible for them to make sure that at no point,

even if our web server is compromised by some adversary,

can that adversary see the submissions.

I definitely think that sources

who know how to use GnuPG already

would definitely value that they can,

on their own end, locally encrypt things,

and that that is secured all the way

to an air gap in our model.

And I think that might make the difference between some sources

who are aware of what these technologies do

feeling comfortable submitting these documents.

For more infomation >> GnuPG Stories: Noah Vesely from SecureDrop - Duration: 3:56.

-------------------------------------------

"Harris J Edits (Mera Saiyaan Superstar)"(Türkçe Çeviri) - Duration: 1:49.

For more infomation >> "Harris J Edits (Mera Saiyaan Superstar)"(Türkçe Çeviri) - Duration: 1:49.

-------------------------------------------

Wisin, Daddy Yankee, Pitbull, J Balvin, Anitta, Maluma, CNCO, Y Mas! Fiesta Latina 2017 - Duration: 1:00:55.

Don't forget to SUBSCRIBE, like, comment and share the mix if you enjoy it!

For more infomation >> Wisin, Daddy Yankee, Pitbull, J Balvin, Anitta, Maluma, CNCO, Y Mas! Fiesta Latina 2017 - Duration: 1:00:55.

-------------------------------------------

Maluma ✘ Shakira ✘ CNCO ✘ Daddy Yankee ✘ Wisin ✘ J Balvin ✘ Nicky Jam - Estrenos 2017 Reggaeton Mix - Duration: 57:21.

Don't forget to SUBSCRIBE, like, comment and share the mix if you enjoy it!

For more infomation >> Maluma ✘ Shakira ✘ CNCO ✘ Daddy Yankee ✘ Wisin ✘ J Balvin ✘ Nicky Jam - Estrenos 2017 Reggaeton Mix - Duration: 57:21.

-------------------------------------------

Liouville's number, the easiest transcendental and its clones - Duration: 20:40.

Welcome to another Mathologer video.

Liouville's number the monster up there

consists of infinitely many isolated

islands of 1s at the 1! th, 2! th,

3! th, etc. digits with

exploding gaps of zeros between them. As

I promised you at the end of the last

video, today's mission is to show you a

nice visual way of seeing that this

number is transcendental. I'm pretty

convinced that as far as transcendence proofs

for specific numbers goes, what I've put

together here is as simple and as

accessible as it will ever get. Still, if

you make it to the end and understand

all my arguments I think you can be

pretty proud of yourself since not even

that many professional mathematicians

know any transcendence proofs. And, if you

make it to the end, I've got a special

treat for you. I'll show you how you can

use Liouville's number as a template to make

a clone of the real numbers within the

real numbers. This clone is made up of

transcendental numbers like this strange

cousin of Pi, but perhaps the craziest

thing about this clone is that although

it is as large at the set of real

numbers itself, it is of measure 0. This

means that in a sense it takes up no

space within the real numbers. For those

of you in the know this set is closely

related to Cantor sets.

What's a transcendental number again?

Well for a detailed introduction just

watch the last video. So here just real

quick: the transcendental numbers are

defined in terms of what they are NOT.

They are NOT among the real numbers that

you encounter when you're doing

classical algebra, like the integers, the

rational numbers, any of the real numbers

that can be written as rooty expressions

like these, or any of the real numbers

that pop up as solutions to polynomial

equations with integer coefficients like

these. This last set of numbers is called the

algebraic numbers and includes all the

other types of numbers that I mentioned

before: integers, rationals and rooty

numbers. Anyway, we call a real number

transcendental if it is not algebraic,

if it is not a solution of a

polynomial equation with integer

coefficients. Liouville's number or Liouville's

constant was shown to be transcendental

in 1851 by the great French

mathematician Joseph Liouville. It was one of

the first numbers shown to be

transcendental and Liouville's proof of this fact

is probably the easiest transcendence

proof for any specific number. Liouville's

proof is accessible to anybody who has

had some exposure to proofs at

university level real analysis and I have

concluded links to the original French

paper and a modern version of the proof

in English in the description. But since

many of you would struggle with any of

the proofs in these papers or in

textbooks, I've tried to come up with an

alternative way of seeing the

transcendence of Liouville's number that can be

understood and hopefully also enjoyed by

anybody who watches these sorts of maths

videos, including all Mathologer fans who

are still in high school. I'm still

working on the primary school version

but and I am probably never going to get there.

Okay, Liouville's number, lots and lots of

0s and 1s and the 1s are the 1! th,

2! th, 3! th, etc.

digits. Just a quick reminder, 1!

is equal to 1, 2! is 1 times 2 is

equal to 2, 3! is 1 times 2

times 3 is equal to 6, 4! is 24.

Then 120, 720, 5040, etc. a sequence of

numbers that grows rapidly which means

that the stretches of zeros between

consecutive 1s get longer and longer

extremely quickly. Okay, here then is my

proof that Liouville's number is

transcendental. Let me know in the

comments whether this proof works for

you. When dealing with a complicated

number like pi

approximate value often suffice or are at

least a good place to start. To get such

an approximation we often chop off the

decimal expansion at some point.

Now let's approximate the number pi

squared with the squares of these

truncations.

Then these approximations for pi squared

will be spot-on

to a certain digit and may be good enough

for some application we have in mind.

However from some point on these

approximations or go off target. For

example, the last approximation down

there coincides with pi squared in only

the first four digits and the remaining

digits that you can see here are wrong.

Liouville's number L behaves quite

different in this respect. But to start

with the only distinguishable

truncations are those that cut off

after the 1s and these are the first

four of these truncations of L. Now

squaring gives these numbers here and in

contrast to what we had before all the

digits of the squared truncations seem to

be correct. And looks are not deceiving,

this is actually always true and the

reason for this is not too hard to pin

down. It's because all the nonzero digits

of Liouville's

number are spaced further and further

apart and those further down the line

simply don't play any role in producing

the earlier digits of the square of our

number. For the moment I'll gloss over the

technical details but I'll return to

them after I've finished

outlining the proof. Anyway, what happens

if you raise L to a different power, say

5? Well then, as you can see, while the

approximations are still definitely damn

good, things are no longer spot on

for all digits. Here, for example, one digit

is wrong and here three but if you look

closely you'll find that the next

approximation is spot on again, all

digits are correct. And actually things

will be spot on from this truncation on.

So, all digits of truncation L_5^5, L_6^5,

and so on will be correct. In fact,

something similar will be true for all

powers. S,o as we've seen, all squared

truncations are spot on. The same turns

out to be the case for the cubed ones,

fourth powers too, fifth powers we've

already seen that all digits are spot-on

from a certain truncation on. Same thing

for sixth powers, except this happens a

little bit later, and so on, for all

powers. Now,

if instead of just power we look at any polynomial with positive

integer coefficients: all pluses no minuses up there, right? So, if instead of just powers

of the number L we're approximating a polynomial like this, evaluated at the

number L, then we find that all the digits of our approximations will be

correct from some truncation onwards. In particular, in the case of this

polynomial all digits will be correct from L_2 onward. We're getting pretty

close to the punchline. Remember what we have to demonstrate is that L does not

solve any equation like this. To get the proof going we first move all the

negative terms to the right side. Now this results in two of those special

polynomials that only have positive terms, one on the left and one on the

right side of the equals sign. Now let's assume that L actually is a solution of

this polynomial equation. This means that the number on the left of the equals sign

is actually equal to the number on the right. I'll call this number the number

at the top. Let's see what numbers we get on the left and right when we replace L

by one of its truncations. Let's start with L_1 and work our way up. Obviously

the higher the truncation the better approximations for the number at the top

will get on the left and on the right but we also know that in addition to

this all digits of these approximations will be correct from some truncation on.

For the sake of argument, let's say that this is the case from L_5 onward. What

this means is that now the first digit of the approximation on the left is equal

to the first digit of the approximation on the right, the second digit on the

left is equal to the second digit on the right, and so on, for a while, until.... Well

you'd expect one of the two sides to run out of nonzero digits and the other side

to keep on going before it stops. Alright and maybe that's exactly what will

happen. However, surprisingly, and we'll see why

when we go all technical, at this point our assumption that L

solves the equation also implies that, starting with a possibly higher

truncation, let's say L_6, we will always run out of nonzero digits at the same

time in both approximations. So this means that our two approximations are in

fact equal from this truncation onward and this means that just like L itself L_6

and all the following infinitely many truncations are also solutions to our

polynomial equation. But infinitely many different solutions are impossible

because a polynomial equation can only have as many solutions as its degree. For

example, our degree 6 polynomial can have at most 6 solutions and so the

assumption that our number L solves a polynomial equation with integer

coefficients leads to the absolutely impossible conclusion that this

polynomial has infinitely many solutions. And this means that our assumption that

L solves an equation like this was wrong in the first place and that means that L

is a transcendental number. Very neat proof by contradiction, right? Of course

what I did not show you is that all the approximations will eventually be spot on

no matter what polynomial equation we're dealing with. Okay, let's have a closer

look. The different 1s in our number L

stand for different powers of 10. For example, the first one stands for 10

to the minus 1, the second for 10 to the minus 2, the third for 10 to the minus 6,

and so on. And the number L is just the sum of these powers of 10. Now we started by

pondering L squared so let's have a look at how we would calculate this number.

Actually before we do that, let's calculate the square of the second

truncation. Okay, so here we have to multiply every term at the top with

every one at the bottom and then add up. So 10 to the minus 1 times 10 to the

minus 1 is 10 to the minus 2, 10 to the minus 1 times 10 to the minus 2 is 10 to

the minus 3. This times that that, 10 to the minus 3 again. This times

that 10 to the minus 4. Add up everything to get the square we are after. For L squared

we have to multiply again every term at the top with every one at the bottom and

then add up. Hmm, might take a while, right? So 10 ^ -1

times 10 ^ -1 is 10 ^ -2, minus 1 minus 2 minus 3,

minus 1 minus 6 minus 7, and so on. Minus 2 minus 1 minus 3, minus 2 minus 2 minus

4, and so on. And so on. Now why are the digits of the square truncation correct?

Well, its rightmost digit 1 comes from multiplying the rightmost term of the

truncation with itself. For the other terms to interfere with the digits of

the square truncation, one of the terms outside the box at the top multiplied

with one of the terms at the bottom should be at least as big as

10 ^ -4. But the largest number we can make this way is 10 ^ -6 times

10 ^ -1 which is 10 ^ -7, a lot smaller than e

10 ^ -4. Okay so there's no danger here. What about these critical numbers for

the next truncation. Also no danger and to gap between the exponents minus 12

and minus 25 is larger than before, so even less danger. One more. The gap has

increased even further. Now it's a one-liner to show that this trend

continues and I actually leave it to you to sort out the details, and, well,

actually you only get partial credit if it takes you more than one line. Let's

have a look at all this for the fourth truncation of the fifth power of L.

That's the next thing we looked at, right? As I said before, all digits are correct

for this fourth truncation. Let's double check this: yes, the green 124 is larger

than the yellow 120. Tick! Again, I leave it to you to fill in the details

to show that this will also be the case for the higher truncations. Remember that

there were problems with the lower trunations. Let's see where these show up

here. Okay the order has flipped, the 30 is now

larger than the 28 and so there will be wrong digits in the approximation. One

step down there's also trouble with the green terms interfering with the digits

of the power of the second truncation. Okay so here are the first digits of L

squared. I just like to highlight again the gap between the yellow lowest terms

of the squared truncations together with the corresponding worst-case-scenario

green terms. As we've already observed this gap is increasing. In fact, it will get

arbitrarily large. And the same is true for all the other powers once the gap

starts appearing from some truncation onward. For those of you who did those

one-line calculations earlier on will already have convinced yourself of this

fact. What about polynomials with positive integer coefficients. Well,

they're built from powers. Take, for example, L squared plus L. Highlight the

second truncation. First note that the corresponding green digits are basically

aligned. That's because the top one essentially comes about from multiplying

the bottom one by 10 ^ -1 that's a shift of one position. Anyway

this relative alignment will always be the same as we move to the right. Also, as

we move towards the right, the gap of zeros in front of the green digits gets

as large as you wish. Given this gap of zeros, it is clear that when we add up L_3

and L_3 squared none of the green digits or beyond will

add to the sum within L squared, and so the digits of the approximation will all

be correct. But what happens if the individual powers in the polynomial are

multiplied by non-trivial positive constants. For example, let's multiply the linear term

by something huge, say 10. The overall effect of this changed term are it's

nonzero islands possibly growing and the greens shifting to

the left. This may lead to the overall gaps in front of the greens shrinking or even

vanishing for really large coefficients like googleplex however since the gaps

grow unlimited in size they will eventually absorb the shrinkage and gaps

will always be present from a certain point on. And this means that no matter

what polynomial we're dealing with, all our approximations will be spot-on from

a certain truncation onward. Almost there. In the final part of the proof we assumed

that L solves a polynomial equation like the one up there. I claim that this

implies that from a certain point onward all truncations also solve this equation

which then gives the contradiction. Well, as just now, we can see that from a

certain truncation, say L_5 onward, we have the usual common gap of zeros for all

the terms in this equation. But then for the identity on the left to be true,

which is our assumption, the sum of the two approximations on the top must be

equal to the sum of the two at the bottom, which is exactly what we needed

to convince ourselves of. Anyway that's basically it. Of course I could still

fill in all the nitty-gritty calculations, for example about the ever

growing gaps but if you've made it up to here I'm sure you'll be able to fill in

those details yourself. Here are a couple of super cool facts that can be shown in

the same way. First even when interpreted as base 2, 3, 4, etc., Liouville's

monster will always be a transcendental number. Second even when we replace the

ones by other digits we get a transcendental number as long as

infinitely many of these digits are nonzero. Now adding an integer to a

transcendental number gives a number that is still transcendental. Maybe

someone prove this in the comments. So adding 5 to the transcendental

number up there give a new transcendental number: five point one two zero,

zero, zero, and so on. Here's a fun idea then. Let's take a real

number like pi and use its digits to create a new number like this. Well

it is super-tough to prove that pi is transcendental whereas proving the

transcendence of this weird clone is not any harder then what I've showed you

earlier. Ok, next trick. Let's do the same for

every real number. What this does is create a clone of the set of real

numbers within the set of real numbers. It consists entirely of transcendental

numbers. Even the clones of the algebraic numbers are transcendental. We just have

to be a little bit careful here when we translate a number that features a

terminating decimal expansion like for example the number 1.23.

Using that terminating decimal expansion would also result in a clone with a

terminating decimal expansion and any number like this is rational and

therefore not transcendental. Of course that is easily fixed because every number

with a terminating decimal expansion has a second decimal expansion with an

infinite tail of nines for example 1.23 is equal to

1.2999..., and so on. Using this alternative decimal

expansion which features infinitely many nonzero digits will then give a

transcendental clone. Well one mini problem remains: the number 0 cannot

be massaged in this way to give a transcendental number. But, okay, as it

happens, quite often 0 is a little exception here, big deal. Anyway, we now

have got a clone of the real numbers within the real numbers consisting of

transcendental numbers. Of course, since it is a clone, there is a one-to-one

correspondence between it and the original set. In other words, just like

the real numbers the clone is an uncountably infinite set. But, of course,

in addition all its elements are easy transcendentals super cool, right? However,

this uncountably infinite set also has the paradoxical property of having

measure zero. So although it is as large as the whole set of real numbers, it is

so well hidden within that in a sense it's not even there. To really be able to

appreciate this paradoxical set you should watch the previous video where I

also give a little intro to how you can assign a length or measure to a subset

of the real numbers, or you could head over to Infinite

Series where Kelsey is also discussing these sorts of sets at the moment. I've

also prepared a proof of the fact that the clone has measured zero but I think

this video is already getting quite long and so I'll put this proof on Mathologer 2

sometime in the next couple of days. Okay, let's finish here. I hope you all

understood and enjoyed this video. But, as usual, please let me know what worked for

you and what didn't and if you are struggling with anything just ask. And

that's it for today. [Music]

For more infomation >> Liouville's number, the easiest transcendental and its clones - Duration: 20:40.

-------------------------------------------

Diffusion-Tensor Imaging of the Physes (July 2017) - Duration: 13:42.

[music]

Ali Guermazi, MD, PhD Good morning. Today we are going to talk about the paper that is going to be in the July issue

of Radiology titled "Diffusion-Tensor Imaging of the Physes: A Possible Biomarker for Skeletal Growth——Experience with 151 Children."

I have the immense, the great pleasure to have the corresponding author with me today,

Professor Diego Jaramillo from the Miami Children's Hospital.

He is an Associate Chair and Professor of Radiology there.

So thank you Diego for being with us this morning and my first question would be can

you explain what prompt you to conduct this study?

Diego Jaramillo, MD, MPH This study is actually the culmination of a number of years of work.

Initially we started working on animal models to look at the tensors in the physes and it

seemed promising so we started taking some – because we were doing some diffusion studies

on regular patients so we would then do the diffusion tensor images and process them in

the physes and we noticed that you could actually look at the columnar architecture of the physes

and the good thing of this is that with conventional MRI you see the physes as a series of layers,

but functionally the physes is really a series of columns and that is what the diffusion

tensor imaging information provides us with.

We thought that we could then start looking at children and we noticed first that, and

this was in a study published in Radiology in 2014, we noticed that the tensors were

longer in the femur than in the tibia and this is what we would have expected if they

would reflect rate of growth because femoral growth is greater than tibial growth, and

then within the femur the growth was greater in the periphery than in the center which

is what we would have expected as well.

That made us think that the modality could be used as a biomarker for growth.

So then we took a cohort of patients that were having MRIs for other reasons.

We excluded any patient who had pathology and then we performed diffusion tensor imaging

on these patients and then recorded several parameters age, gender, BMI, etc. and then

we noticed that there was a striking correlation between growth or expected velocity of growth

in a number of the DTI parameters.

This is actually amazing because you started in 2014 it looks like with animal model and

the discussion of this study that you published and can you just explain in more detail what

the findings are, like briefly and why they are important?

The length has been shown histologically that rate of growth depends on the length of the

columns of converse lengths in the physes.

We noticed that the tensors not only extend into the physes but also into the metaphysis

in the newly formed bone.

But still the column of cartilage in newly formed bone corresponds to the rate of growth.

And so when you start looking at DTI parameters to make the length of the tract or particularly

the volume of the tracks, they correlate well with the expected growth.

So then we looked at the patients and we noticed that in our population the girls started having

longer and longer tracts as they approached puberty and so by age 11 they had a peak and

then the peak came down.

Boys had the same findings but two years later.

These are the same findings that we see for example in the Tanner growth curve where the

greatest speed of growth is in girls around 11 and in boys around 13.

And then we looked at other parameters, for example ADC.

Interestingly ADC does the same thing, but a couple years earlier.

But then we also looked at short patients versus tall patients that were greater than

the 50th percentile.

Of course the tall patients had longer, and this was particularly true in girls, the tall

girls had longer tensors and larger tensor volumes than the short girls.

We haven't done a longitudinal study but it seems to be related to speed of growth.

You said it well.

It looks like this is adding a layer to what the x-ray can inform us about bone growth,

so what is then the real add value of your described method?

The real added value is that you could have a biomarker for growth and have more information

about two things; how much growth there is going to be, and secondly is there physes

getting too close to physeal closure so is there no more residual growth left.

The importance of this is for example with therapies like growth hormone where we administer

growth hormone in patients and it takes about a year or sometimes two years to determine

whether there is response to therapy.

Growth hormone costs approximately $50,000 per year.

If you could identify the subset of patients that are responding to therapy after a few

months, then you could decrease potential complications from therapy and also save a

lot of money.

This is absolutely great idea.

I do think as an application for growth problems is probably one of the most challenging problems

we have today and having a biomarker like this one would be absolutely fantastic.

Then it comes actually to the next question which is obviously if you have this idea about

implementing in treatment in patients with growth problems, do you think this method

can be easily implemented everywhere or do you think the method as any others comes with

some challenges and if there are challenges, can you just tell us about them?

We have used off the shelf products, so this is the diffusion weighted sequence, we used

– all the imaging was done at three tesla on Siemens three tesla units and using the

conventional diffusion sequence that you can get on the magnet with 20 directions and b

values of 0 and 600, so it's something that is readily obtainable really in any of the

MRI units.

Incidentally we did obtain similar measurements at 1.5.

So it's not exclusive to 3 tesla imaging, you can get pretty good imaging at 1.5 as

well and also and then the reconstruction is done with software called Trackvis, but

you can use other software to reconstruct the tensors, but again I mean this is something

that is readily available.

The tricky part is when developing the expertise to sample the right area of the physes, but

this is something that I mean several on our team have become very proficient and it's

– I mean it's not immediately easy, but it's not terribly difficult either.

I mean it's not that more difficult than performing 3D reconstructions for example.

Okay so if I understand you well you use it in specific situation at this point?

Yeah so right now we're conducting a study, an interesting study, with survivors of neuroblastoma,

and these are patients who have a lot of growth problems afterwards and many of them require

growth hormone administration.

Interestingly, we've seen that the for example the fractional anisotropy is very different

for the neuroblastoma patients than for the controls in that the patients who have less

growth have shorter tensors, a more disorganized growth plate and they don't have as a high

NADT so all of these parameters that come from the DTI seems to be correlating well

with the rate of growth in the patients who are receiving growth hormone and that are

having growth problems in the case of neuroblastoma.

The plan, a little bit longer term, is to follow a cohort of patients who are short

and who some of which have growth hormone deficiency and some of which don't, and

also to do a longitudinal study in normal patients because one of the things that we

don't have is longitudinal data.

We don't have – we have cross sectional data from a lot of patients, but we haven't

followed patients let's say for two or three years and that's another thing that we plan

to do.

Very interesting and I'm looking forward to the next data longitudinally.

I hope that you're going to just think about Radiology and to send it there.

Of course.

Looking forward to it.

I have my actually, I do think we are in the last question and I really, I mean I'm actually

have been doing some pediatric radiology in the past, MSK, and to me is actually the bone

age is a big issue that is very important.

So do you think we can use, with confidence, this method in resolving bone age issues or

simply your method just focuses on skeletal maturation?

I think at this point we don't have enough data to have that degree of reliability.

I think one thing that we can predict with a lot of certainty is whether there is growth

potential or not, but beyond that in terms of dating the growth plate or the skeleton,

I don't think we're quite there.

One thing I was going to say is that another application going forward will be for example

in patients who have chronic injuries to their growth plate, the thing that comes to mind

which his very common is patients who have suffered chronic physeal injury due to sports;

gymnasts, baseball players, etc.

Again in these patients it's very important to know, should they stop their activity or

is there a damage to the growth plate or can they continue exercising.

Again, that's another area where we would like to explore.

But in all these cases I don't think we have enough data yet to really know with certainty

for example what is going to be the growth in the next few years and so the only thing

we can say with certainty is there growth potential or not.

Thanks Dr. Jamamillo, thanks Diego.

It's been a pleasure talking to you and thanks very much for joining from Switzerland.

This is tremendous work and it's very exciting and I'm really looking forward to some longitudinal

data that we're going to receive hopefully in Radiology.

Thanks again.

Thank you.

For more infomation >> Diffusion-Tensor Imaging of the Physes (July 2017) - Duration: 13:42.

-------------------------------------------

Diffusion-weighted MR Imaging for Prostate Cancer Detection (July 2017) - Duration: 17:49.

[music]

Herbert Y. Kressel, MD Hi. This is Herb Kressel and welcome to the Radiology podcast.

Today I'm joined by Dr. Andrew Rosenkrantz, Associate Professor of Radiology and Urology

at the New York University School of Medicine.

Dr. Rosenkrantz and his colleagues authored a fairly definitive paper entitled "Diffusion-weighted

MR Imaging for Detection and Assessment of the Aggressiveness of Prostate Cancer: Comparison

between Conventional and Kurtosis Models."

Welcome Dr. Rosenkrantz.

Thanks so much for joining us.

Andrew Rosenkrantz, MD Thanks so much for having me.

Sure.

Many of our readers may only be peripherally aware of kurtosis, diffusion kurtosis imaging,

and its impact in assessing the aggressiveness of prostate cancer in particular, so could

you tell our listeners and viewers what actually is diffusion kurtosis?

Okay so I would compare diffusion kurtosis imaging to conventional diffusion-weighted

imaging which assumes that the water diffusion is free in its behavior.

Diffusion kurtosis imaging recognizes a broader range of hindrances to water diffusion and

can account for those hindrances in its modeling to extract information about the microstructural

complexity of tissue.

It can look at how complex the tissue is at the cellular and even the intracellular level.

So the more complexity of the tissue the less free the diffusion is and that would affect

the kurtosis.

It also affects the distribution of the diffusion parameters.

Correct.

Okay.

With that as a basis, why would you think this kind of a metric might be of value in

assessing the aggressiveness of prostate cancer in particular?

So for assessing the aggressiveness, prostate cancer we're really talking about the Gleason

score which currently it's a score that the pathologist determines on their visual

assessment looking at the tissue under the microscope.

That is really driven by the architecture, the complexity of the tissue in its appearance.

The conventional ADC is kind of more of this hodge- podge or amalgamation of facts also

influenced by tissue viscosity, other factors, and the hypothesis is that the diffusion kurtosis

is more of a kind of a pure or less contaminated reflection of really just the structural complexity

and hindrances which its believed could more reflect what pathologists are trying to estimate

when they assign the Gleason categories.

Thank you.

Now this study is a follow-up to your own which we published in 2012 was a preliminary

assessment of diffusion kurtosis imaging in prostate cancer, and since that one was published

there have been several others published and the results as you note have been mixed.

Why did you think that this current study you're reporting could more definitively

determine the utility of kurtosis in looking at the aggressiveness of prostate cancer?

Were there particular problems with the prior studies?

Yeah, so the studies I've been – so this has been a topic of a lot of interest in the

last few years, we're all trying to best predict the Gleason score non-invasively which

has huge implications for prognosis and treatment selection and the studies have had mixed results.

Some have been really promising, others have not been.

So in terms of some of the limitations of what's been done is the sample sizes have

generally been small.

There were some use of the reference standard biopsy instead of radical prostatectomy as

we did in our initial feasibility study of kurtosis in the prostate which has diffusing

biopsy has limitations in terms of the precision of the grading, and even just some technical

factors in terms of how the kurtosis image sets are processed and the metrics are obtained.

I think some of our understandings of this have evolved and improved over the last few

years, and I think in terms of what b values from a diffusion acquisition are incorporated

and how this is done.

It's not just pushing a button.

There's a lot of optimization of the (inaudible.)I think even that has gotten more precise as

well.

Good.

Can you briefly outline what you actually did in your study?

Tell us a little bit about the design and sort of what you looked at.

We had a sample, a retrospective sample, of several hundred patients.

At our institution our standard diffusion acquisition provides the necessary information

to be able to extract the kurtosis metrics.

All the patients underwent prostatectomy after MRI.

We included patients where pathology had done detailed mapping of the distribution of tumor

within the prostate including localization and grading of individual tumor foci.

We then selected the dominant or index tumor in each patient, and correlated these back

to the imaging and placed regions of interest to obtain what we deemed to be the optimal

ADC, standard ADC, and the optimal kurtosis metric that we could generate using our acquisitions.

Were these mean ADC and mean kurtosis over the whole volume?

It was the mean value of an ROI on the higher lesion.

And then we correlated that back to the pathology and we looked at the ability of ADC and kurtosis

to separate benign versus malignant as well as to separate low grade from high grade tumors

using two different definitions of high grade that have been applied in the literature.

So really we had three different histologic outcomes for which we compared conventional

ADC and kurtosis.

Okay and what were your key findings?

In terms of our findings, we found these interesting.

ADC and kurtosis they were highly correlated.

As ADC goes down, kurtosis goes up.

Both of those show as being generally associated with greater aggressiveness.

Their accuracy for predicting our various end points, benign versus malignant or high

grade tumor, were very similar.

When we looked at the subset of patients let's say there was a difference between them, kurtosis

actually did not outperform ADC within that subset.

So either they generally gave – for most patients they predicted the same thing.

Either ADC and a kurtosis said could be benign or be high grade or both of them said there'd

be, they both gave similar results or basically giving – to say it somebody's giving some

redundant or repetitive information from the kurtosis that we already have from ADC and

in the small minority when they were discrepant, the ADC actually did similar or even a little

better.

Although in the paper you describe what I would consider a fairly rigorous protocol

for matching the images particularly the target lesion with histopathology.

In 36 patients the target lesion couldn't actually be identified on the MR images.

Were you surprised by this and why do you think it might have happened?

So that's a really important point.

Those were overwhelmingly low grade tumors.

These are Gleason 6 lesions.

This is recognized actually in prostate cancer, this is a recurring issue whenever you try

to do detailed MRI pathology correlation for prostate cancer.

And other groups have described, other (inaudible) have described this as well.

Some prostate tumors they don't really form a discrete mass like say a renal cell carcinoma,

it's just it's histological sparse.

The malignant glands are interspersed with benign glands.

There's really no actual discrete tumor and these can actually just be invisible on

imaging.

For those we had no clear correlate on the imaging.

And for those in our final analysis, we ended up basically looking at the numbers both ways

where we just empirically placed an ROI and where is seemed the lesion was based on pathology

versus just excluding those.

Whether we did an empiric ROI or we left those out we had similar.

Our final observations held true.

Okay.

Now although the ADC and the kurtosis findings were correlated and concordant and around

80%, as you noted they were discordant at around 20% and you mention that in these the

majority of them the ADC was sort of more accurate or at least correlated better with

the histopathology, but my question is what do think is driving the discordance and what

sorts of lesions did you notice this discordance?

ADC and kurtosis they were correlated but these aren't – there are differences.

The ADC does have influences from viscosity concentration.

There are some other – it is kind of more this kind of this messier metric whereas the

kurtosis is really more tuned into complexity and hindrances to water motion.

There will be a small fraction where they didn't exactly follow each other.

I can't say exactly what the unifying kind of property was of the lesions where that

happened, but that it was not common and the ADC was either similar or a little better

than kurtosis in that.

So what you're saying is that they are in fact, although they're related, they're

measuring somewhat different things and it's not surprising that the measurements wouldn't

be the same in some subset of patients.

Exactly.

Now as we said in your study you looked at whole ROI or most of the lesion volume ADC

and kurtosis.

Others have started to look at subsets doing histograms of the ADC and the kurtosis looking

at sort of the worst actors, the bottom 10%, bottom 25% and have you looked at other kurtosis

based metrics?

So I think the histogram metrics for extracting a summary value from diffusion acquisitions

is really promising.

So say looking at the 10th percentile or (inaudible) this has shown I mean some really compelling

results in the prostate as well as in a range of other organs where work from other centers

or work that we've done at NYU.

I think unlike the diffusion kurtosis technique this doesn't really require any changes

or adjustment in the protocol or a standard protocol can be done and it's all on the

post processing side.

I think vendors are making those tools available and this is I think becoming easier to incorporate

into clinical workflows.

So I think this has a lot of potential in the near term.

The same thing with the kurtosis method or a histogram of it?

And this can be completely applied to the kurtosis data set which we didn't do in

this study, but I mean it's a great idea.

The literature has consistently shown incremental improvements in performance when applying

the histogram analysis.

So whether done to a conventional ADC, you could do a conventional diffusion analysis,

it could be applied in a comparable way to the kurtosis data set.

It wouldn't surprise me if that led as well to incremental improvement there.

In this study we were just doing a comparison of conventional mean between kurtosis and

standard diffusion, but I think I would anticipate that applying a histogram approach to either

of these could bump up their performance a bit.

Yeah I mean histologically as you know the Gleason grade can vary throughout a lesion

in scattered areas and so it would seem that sort of you know the and my understanding

is that the worst grade drives the final grade so that this might more closely mirror what

the pathologist is looking for and sort of finding the most aggressive areas to establish

the grade.

I agree with you, it sort of makes sense as an area for future investigation.

This sort of leads up to my final set of questions for you, it sounds like you haven't given

up on kurtosis yet.

Are you still acquiring this data in your patients?

We're looking at, I think more can be done to optimize it and there might be some additional

ways of, some additional angles that haven't really been as well explored or considered.

So no we haven't given up on it.

I think it can be something that can be combined with other innovations.

As an example, more recently what we're turning our attention to is time dependent

diffusion.

This is a parameter that hasn't gotten quite as much attention by radiology investigators.

A lot of the focus has been on the b value and directions of the underlying diffusion

acquisitions.

There's also the actual diffusion time.

The time between the diffusion (inaudible) gradients actually can be manipulated.

It's playing with the actual source sequence.

The vendors don't always make as readily available to us as radiologists to work with,

but it can be, it actually can be adjusted and we're coming back and actually – we

can actually push it to really short diffusion times, shorter than are usually applied and

can technically we tune the contrast of the diffusion image sets in that manner, we can

actually go back and combine that with the kurtosis analysis or the conventional diffusion

analysis; and then we're finding some really interesting results that way.

So we're still using kurtosis but the idea is to combine this with some other manipulations

of the diffusion acquisition and we'll see what it shows over time.

Sounds like there may be another Radiology podcast in your future as you get more into

this work.

Dr. Rosenkrantz it's been a pleasure chatting with you about this paper, and I think at

this point in time it certainly gives us a good handle on the relative role of diffusion

kurtosis imaging in looking at prostate cancer aggressiveness.

Thanks so much for joining us.

Thanks so much again and I really enjoyed participating.

Great, bye-bye.

Bye.

For more infomation >> Diffusion-weighted MR Imaging for Prostate Cancer Detection (July 2017) - Duration: 17:49.

-------------------------------------------

Ford S-Max 1.8 TDCI Navigatie 7-Persoons Trekhaak - Duration: 0:55.

For more infomation >> Ford S-Max 1.8 TDCI Navigatie 7-Persoons Trekhaak - Duration: 0:55.

-------------------------------------------

NFL Star Racks Up $102K Bar Tab - Duration: 0:40.

DON'T BELIEVE EVERYTHING

YOU READ ESPECIALLY FROM

A THIRD-PARTY

ENTERTAINMENT NEWS

SOURCE TRYING TO GET

CLICKS.

NFL STAR ROBERT RON

KAUTSKY SURENESS HAD A

PARTY.

JEFF: THE PATRIOTS TIGHT AND

RACKED UP 102,000-DOLLAR BAR TAB

AT A CASINO IN CONNECTICUT.

IT WENT DOWN AT THE

SHRINE AT NIGHT CLUB.

HE WAS SEEN TEARING IT

UP ON THE DANCE FLOOR AS

HIS PALS RAN THE EPIC

BAR TAB UP.

$20000 WORTH OF

DOMPERIDONE ROSÉ, $17000

WORTH OF ACE SPADES ROSÉ

AND ANOTHER 10 GRAND OF

For more infomation >> NFL Star Racks Up $102K Bar Tab - Duration: 0:40.

-------------------------------------------

Sirf 2 secnd Ka Wazifa in Urdu | Sirf 1 baar parhian by Kamran sultan - Duration: 2:01.

For more infomation >> Sirf 2 secnd Ka Wazifa in Urdu | Sirf 1 baar parhian by Kamran sultan - Duration: 2:01.

-------------------------------------------

The Case For Eating Bugs - Duration: 4:29.

Would you eat bugs if it meant helping to save the planet?

I'm Anna and this is Gross Science.

Let me start off by saying that, depending on where you live, the idea of eating insects

may not gross you out at all.

According to a 2013 United Nations report, insects are included in the traditional diets

of at least 2 billion people.

That said, most Westerners, especially in northern latitudes, don't tend to eat bugs—and

in fact, find it pretty disgusting.

So, why is that?

Well, no one knows for sure, but there are a bunch of ideas.

One is that it's simply easier to eat bugs in the tropics, because they tend to be larger,

there's a diversity of species available there all year round, and those bugs may have

predictable harvesting seasons.

Another idea is that traditional agriculture may have made eating insects less desirable.

Agriculture first began in the Fertile Crescent, and then spread to Europe.

And those regions already had large mammals that could be domesticated, so collecting

bugs for food may not have been as much of a priority.

And, as time went on, insects may have gone from being accepted as benign parts of the

environment to being seen as agricultural pests, and invaders in our homes.

Whatever the reason, the fact that most people in the US and Europe regard the idea of eating

bugs with disgust is mostly cultural.

After all, when you think about it, it is a bit strange that our mouths water for this

creature with an exoskeleton and pinching claws and not this one.

Now, you might be saying to yourself, "Who cares!

I still don't wanna have bugs for dinner!"

And to that I say, "Let me lay out the case for entomophagy—the practice of eating insects."

To begin with, insects are efficient at putting on weight.

In order to raise a cow to maturity, you need a lot of food.

But to rear the same amount of cricket meat takes 12 times less feed.

What insects eat is also more sustainable.

We need to devote a lot of agricultural land to traditional livestock—so we can grow

food for them, or so they can graze.

But many bugs can live off of waste products, like manure and compost.

That said, the risks to us of raising edible insects on organic waste are still being studied.

On top of that, livestock like cows are notorious for releasing tons of methane, which is a

potent greenhouse gas.

But very few insects produce methane at all.

Of course, none of that would matter if bugs were bad for you.

But it turns out that they can be a great source of proteins, healthy fats, and minerals,

depending on the species.

Clearly, there are lots of economic, environmental, and health reasons to eat insects—though

we'd have to figure out sustainable ways of harvesting lots of bugs if entomophagy

were to catch on more widely.

But that's not gonna happen if people don't like how they taste.

So, how do bugs taste?

Well, people say crickets taste nutty, scorpions taste like shrimp, and the grubs of the palm

weevil taste sweet when they're raw or like calamari when they're cooked.

And I can tell you that everything I ate today was pretty good, except for those grasshopper

tacos, which were really not great.

But I think that was just probably because I cooked them poorly.

I don't blame the grasshoppers.

I'm gonna put links in the description to all the recipes I used today, and you can

try them out yourself.

I especially recommend the scorpion curry.

And if you do try bugs, let me know what you think, and how you prepared them!

Take a photo of your creations and tag me on Instagram!

I'd love to see what you cook up.

Ew.

By the way, there's one more reason to eat insects—as pest control!

Palm Weevil grubs (the ones that taste sweet or like calamari) are becoming a big problem

in places like southern California.

My friends over at Deep Look actually made a whole video about these guys.

So head over to their channel to see how weevils could harm the iconic palm trees of Hollywood.

And if you like Gross Science, hit subscribe!

That is not delicious.

For more infomation >> The Case For Eating Bugs - Duration: 4:29.

-------------------------------------------

Boy or Boys - Duration: 0:56.

SNSD - The Boys

SNSD - I Got a Boy

SNSD - You Think

SNSD - Gee

SNSD - Genie

Hyuna - Ice Cream

TWICE - Knock Knock

April - Mayday

WJSN - Mr. BADBOY

EXID - Boy

AOA - Oh Boy

Apink - Drummer Boy

gugudan - Good Boy

AOA - Crazy Boy

B1A4 - You're a Girl, I'm a Boy

EXO - Playboy

Dal Shabet - Ok Boy

Thank You For Watching! Remember to Subscribe! = w =