Hi there, it's Matt here and welcome back to the podcast and welcome back to another of our Ask Me Anything episodes, our AMA episodes.

And as always, that wonderful presence in the AMA show that is Dr. Eity Ben-Simon is back with me making my Dulcett non-intelligible, non-meanful answers,

wonderfully meaningful. Thank you for rejoining the show. Welcome back. Thank you. Lovely to be here.

So let's fire away with the first question.

The first question, I feel, is touching on something that I see everywhere around me lately, which is

AI. Someone is asking, what is the future of AI and machine learning in diagnosing and

treating sleep disorders? Interesting question. Yeah, it is an

interesting question. What's the current sort of applications of AI and ML? I'm going to try and pretend

that, you know, I am an advanced sort of machine learning data scientist. I think AI and machine learning

have certainly begun to impact sleep medicine,

mostly by way of the accuracy and the efficiency with which we do diagnostics. And I think one

significant application here is probably the automated scoring or what we call staging of sleep and sleep studies. We used to have to and still do

in the lab, in fact, we will record people's sleep in the laboratory and then we will scroll

through 30 second by 30 second across their entire night and we look at all of their features

with trained minds that are scientifically

trained to evaluate and score sleep, and we look for all sorts of different things.

It's immensely labor intensive, and there is significant human error.

There have been now some groundbreaking studies, I think, starting in kind of like 2017, 2018, that introduced

deep learning algorithms that were able to accurately classify individual sleep stages, so REM,

non-REM, and the different stages of non-REM, light non-REM, deep non-REM, using just one EEEG

channel on top of the head, and it was able to match

the performance of expert clinicians.

...