Predicting AF 30 mins before it happens

[Barny12]

Researchers at the University of Luxembourg (LCSB) have developed an AI which is able to predict an episode of afib 30 minutes ahead of time with around 80% accuracy.

sciencedaily.com/releases/2...

Prof. Jorge Goncalves, head of the Systems Control group at the LCSB. "We used heart rate data to train a deep learning model that can recognise different phases -- sinus rhythm, pre-atrial fibrillation and atrial fibrillation -- and calculate a "probability of danger" that the patient will have an imminent episode." When approaching atrial fibrillation, the probability increases until it crosses a specific threshold, providing an early warning.

"Another interesting aspect is that our model has a high performance using only R-to-R intervals, basically just heart rate data, that can be acquired from easy-to-wear and affordable pulse signal recorders such as smartwatches. These devices can be used by patients on a daily basis, so our results open possibilities for the development of real-time monitoring and early warnings from comfortable wearable devices,"

The long-term objective is for patients to be able to continuously monitor their cardiac rhythm and receive early warnings that can provide sufficient time to take antiarrhythmic medication or use some targeted treatments to prevent the onset of atrial fibrillation.

Used on a smartwatch could this be a game changer? What do you think?

[Hidden]

cell.com/patterns/fulltext/...

Potentially a major breakthrough, though numbers in the study are low at 350

[jeanjeannie50]

Sounds amazing!

[Afbiff]

sounds great and this made me wonder if a dog can pick up an impending AF event - am I right in thinking dogs can predict an epilepsy event? If so, why not AF?

[OzJames]

sounds promising, they should start a Trial and I’ll join in!

[BrotherThomas]

Perhaps the downside is that the anxiety of knowing that you're going to have an attack is enough to trigger it sooner.

[bean_counter27]

What does that mean for our AF triggers? Does it mean if they are a result of triggers we generally would need to be "exposed" to them more than 30 minutes prior to AF episode - and in cases where the triggers need time to have effect e.g. digest food/drink, even longer? Or do they fall into the 20% they can't predict 30 mins beforehand. Sounds like a lot of research potential on triggers could also ensue.

Or does it mean they are detecting vulnerability to an AF episode but they still need to be triggered, which only occurs ~80% of time. If so, could be a potential warning we could heed to help prevent episode by taking certain actions e.g. need to calm down or rest, hydrate etc

Lots of research potential.

[50568789]

What if they researched at the other end and developed a smart watch that could deliver a little shock to stop the Fibrillation and put you back in sinus? Dream on, I suppose.

[AussieHeart]

I get a warning prior to a migraine with squiggles running beside my eyes. If I take medication the migraine doesn’t develop. If this AI meant I take a PIP (as it takes about 20-30 mins to kick in) and stopped my AF episode I’d be very happy with that!

[Barny12 in reply to AussieHeart]

And if you can prevent pretty much every episode of afib you stop the progression of the condition, too!

[Cavalierrubie]

l don’t think l would want to know this as it would give me anxiety and probably cause lots of problems. Now, if they could come up with a way to stop AF before it even starts, then that would be amazing.

[Barny12 in reply to Cavalierrubie]

But knowing that you could just pop a flecainide and stop the episode before it started would surely ease your mind?

[Cavalierrubie in reply to Barny12]

No no really Barney. I don’t control my AF with Flecainide. I can’t tolerate most of the drugs. Also, it would have to be 100 per cent accurate in its prediction otherwise you are medicating unnecessarily and flecainide is a pretty potent drug and for some it can stop working after frequent use. I see the point you are making, but that may work for some, but not all of us. I still would rather not know if l was going into AF as it would, for me personally, interfere with my life like a black cloud on a summers day. Each to his own.

[108cat]

This has very interesting potential as Beancounter suggests in terms of further research into triggers and how they work. However I'd rather not know AF is on it's way, as Brother Thomas suggests, it could create more anxiety and stress. I suppose for some it might be helpful ...

[Vicmeister]

Think this is amazing. On my apple smartwatch there is the heart rate variability. Just before I have an AFIB attack the variability sky rockets to over 100 and above. At first I didn't know that this high reading was related to an upcoming AFIB attack. But after studying the heart rate variability it is obvious that there is a definite warning when it goes that high. When the AFIB event is over the heart rate variability goes back to normal and stays there every day until the next Afib attack. For me it is between 30 and 40 which according to stats that is healthy for my age. The HR variability goes down with age. I believe that it is an invaluable tool to warn of an upcoming AF event and help with lessening the burden of the attack by taking any medication beforehand.

[ochinee]

" The long-term objective is for patients to be able to continuously monitor their cardiac rhythm and receive early warnings that can provide sufficient time to take antiarrhythmic medication or use some targeted treatments to prevent the onset of atrial fibrillation. This in turn would reduce emergency interventions and improve patient outcomes."

This could be a huge advance in treatment/prevention of AFib episodes!

Thanks for the information!