Tag: AI Clinical Research

The generation of RWE at multisite level using Natural Language Processing on Electronic Medical Records.

NATO summit in Madrid gives you clues on AI models.

These days it is unlikely that your wallet will be stolen in Madrid.

Basically because there is a policeman on every corner.

Even for people like me, who vowed years ago not to watch the news, it’s hard to ignore that there is a NATO mega-summit whose goal is to scare the East.


Or rather to tell it that we are not scared.

Which is not true.

Because we are.


Well, the fact is that this is the way the world works.

The balance between the national and the global.


We are all aware that a single world state in which there were no regional singularities would be gray, dictatorial and inhuman.

And at the same time we have all seen that alliances, union and the abolition of borders make us better.

Balance is tricky.


It’s the same with clinical predictive algorithms and geography.

They have to work at a local level, with a specific population.

And at the same time they have to be usable anywhere in the world.


Look, a couple of years ago, in the middle of COVID, a team from China published in Nature a model made with machine learning that was very good at guessing which cases were going to become severe.
 

And after using it, three groups replicated, from Amsterdam, Paris and New York, saying that they had tried it and it didn’t work for their populations.

So the Chinese team replicated again saying that they had RE-ENGINEERED the model with the data from these places and now it did work with those Western populations.

– There you have it. Try it out and see what nice Areas Under the Curve it is!


Well.

We can call this whatever you want.

Transportability. Calibration. Transfer learning. External validation.

Every time someone rejects a name I invent another one.

Or I take it from the BMJ. Or from JAMA. It’s all there.


If you want to do a predictive algorithm with one population, then you need to transport it to another geography. 

And that has a very precise methodology and technique. With its N and its metrics. It’s not difficult, but you have to be exact and follow the steps well. And if you do it, then it is valid for any place in the world. If you want to see how it is, we can continue here.

Take advantage of the fact that I am writing under propofol.

This pneumologist is several years ahead of the curve and is publishing a large number of papers.

When Dr. Izquierdo, Chief of Pneumology in Guadalajara, gives a talk about his experience with AI, I know I’m going to have a good day.

It’s the easy part of my job.

 

Sitting in the chair and listening to him say:

– I do COPD research.

– And the last study I did in COPD took me years of data collection.

– And by the time I had the data collected it was outdated.

– And then I met Savana

– And now I analyze millions of data in weeks

– And this is all I’ve published

– And this is all I’m going to publish

– Thank you, best regards

 

This is the easy part, I tell you.

 

The hard part is having spent years building tools, inventing methods, validating standards.

Battling trolls, getting around politicians, looking for funding.

Recruiting the best people, designing communication plans, travelling.

 

And above all, getting it wrong.

Because I was number 1 in all of this.

Number one from the bottom.

 

I didn’t understand anything about what something made for research should look like.

So I was making very complicated AI programs, but they didn’t follow the rules of research.

And nobody wanted me.

 

Nobody wanted ME, as good a lover as I am….

 

So I had no choice but to do something very boring, but fundamental.

On the TECHNOLOGY, build a METHODOLOGY.

And everything, but everything, started to change.

 

It is very strong.

 

Because now I have a methodology of such a caliber that it would change the results a lot for anyone who knows how to listen.

I mean, who knows how to listen, not to hear. Listen. Listen and understand.

That is one requirement, and there is another one that almost nobody has.

Pressure.

Time, money or social pressure.

We live in such an affluent society that whoever puts into practice the teachings of this methodology will do so with little or no competition.

Because most of those who come to it will listen to it, like it, dream about it at night and will never do anything I propose.

And that’s fine.

That’s fine, for the few like Dr. Izquierdo, who will, here.

You'll buy me a sea bream to avoid the "resident syndrome".

Look, there’s this thing that Jorge Tello, the co-founder and CEO of Savana, and I call “the resident syndrome”.


I’m going to tell you about it because I think it can help you.


But I tell it to you with a story.



I recently found myself at a talk I went to give to the Supreme Leader of a well-known healthcare group.


I went to greet him and he said something I wasn’t expecting, especially since he doesn’t seem to be a man who is very easy on the ears.


He said “you were the first”.


He meant that he had been the first, it is understood in my country, to see the coming of AI in Medicine.



Then he did not pay much more attention to me and went on with his business.


This despite the fact that we recently applied together for a multimillion-dollar call for applications and were awarded the maximum score.



But the lesson is not in what he told me. It’s that someone who thinks “I was the first” paid little attention to me. I don’t say no attention, because that would be a lie, but little attention.


And I insist that we have things together.


His hospitals participate in our global research network….



But he wouldn’t talk to me for five minutes.


And maybe you’re thinking that I’m bringing all this up out of ego.


But I have to tell you that ego is best thrown down the drain, as soon as you set up a company.


I bring it up because we are dealing with an eminent case of …. (drum roll)… resident syndrome.

 



It’s the one that happens when someone has seen you when you were little.


For him you will always be a resident.


Even if you get Nobel Prize.


It is a price you pay with the first people you work with.


That’s fine, as long as you work.


But you have to know it.



Look…
When we get into a project, you’re going to want us to take responsibility. Of course you will.


But you’re also going to want to have control over what happens.



We don’t want the fact that AI is something new to make us look to the status quo like newcomers who don’t know what they’re talking about.

 


The good news is that at Savana we’ve been given so much crap, we’ve suffered so much from the resident syndrome… that you can now do an AI project with us without the risk of anyone perceiving you as the first year resident.

 

If you want to do a project with just the right amount of responsibility and control, with the authority of a pope, here.

A walk among people who have it bigger.

Yes, I know I told you about the American Conference of Oncology yesterday.

And if you’re not an oncologist, you might think it’s a pain in the ass.


But look, no.

Oncologists are ahead in almost all research.

And what happens there will happen in yours and mine.

That’s why I insist.

Because see, learn this…


In the giant, overwhelming area where the companies’ fair booths are, there was something striking.

There, between the tombola and the cotton candy, there was something there.

Something so big you could miss it.


It was the booths of the Real World Evidence companies.

As bulky and bulky as those of any pharmaceutical company.

And bigger, yes bigger, than those of the clinical trial CROs.


And that, dear contestants, is a sign.

It’s telling you where things are going.

The machine learning thing to do virtual trials.

And the computer-based evidence generation thing.


I went over to talk to some of them.

I like to talk to competitors.

Because I realize that competitors don’t exist.

There is only one’s ability to see the opportunity that opens up, versus the one that closes down.


This is like the beginning of the Internet.

Or like the arrival to Rio Bravo.


They all had exceptional solutions.

Of all of them, I’m sticking with Sophia Genetics.

Nice, smart people.

I highlight them because it makes a lot of sense to increase predictability, putting together genomic data with medical history data.


Look… and I close.

I have only one competitor.

And that is ignorance.

Ignorance of what Real World Evidencce, generated semi-automatically using massive data and machine learning techniques, can do.

If you’ve already figured it out, then you’re ready to talk here.

Enrol in a project