Fundamentals of Using Ai on EEG to Detect Seizures

[@THEMATT222]

Interesting 🤔

[@niyaayin]

Love the efforts👍

[@user-ty2sl6io2p]

Hi! Thank you a lot for your explanation, it was perfect given that there is no much information on the internet. I wanted to ask you why did you choose neural network to came up with this solution. Again, thank you

[@jaredbeckwith]

In the past they used to use hard coded rules, but now programming Ai that learns on human labeled seizures gives better results. Ai is still a long way from being as good as humans on EEG but I thought it was an interesting project!

[@martingallagher9074]

Thanks for posting! What AI software do you use? Why do you lump the FFT data into clinical frequency bands instead of looking at the power at all integer frequencies -- i.e. 1-20? Don't you lose potentially-useful information when you lump it into bands?

[@jaredbeckwith]

Hi Martin, I’m trying to build my own software named AiON. This method was based on some random research paper I was looking at. I never implemented it fully, but it gave me some sort of idea how to detect seizures. I think there’s definitely better methods out there that I want to look into.

[@the_BattleGoose]

What are your thoughts on DiagnaMed Holdings? The Brain Year and Brain Tremor products, do you think they will have a future?

[@jaredbeckwith]

Hey I’m not familiar with their products so I can’t give a good answer. It is somewhat true you can tell how young a brain is from looking at EEG. A healthy young person shows 8-12 waves per second while awake. Once an adult starts getting elderly, you can see slowing of 7 waves per second or less.

[@mybabymxx2944]

just curious if you’re already registered…

[@jaredbeckwith]

Taking test to get registered soon!