 haven't tried by chaos.com yet, please go check out the site. Don't forget to provide your feedback. This tool is actually showing you image compression. It can be used for any image. Of course, in this case, it's showcasing medical data. It can use different... it's a wavelet transform. So you can use different wavelets inside how many bits you have in the image. The level of compression. So essentially, the higher the compression, the more levels, the more compressed it will be. And the PSNR, meant to the peak signal to noise ratio, meant to essentially tell you how easy it will be. You'll be able to reconstruct the compressed image back to the original one. Yes, you can increase the number of levels and increase the number of beads. Yes, we can play around with that. What else we got? The feature extraction. Yes, so it's using the detection type. You have three types at the moment. The fast, high-res sheet, the massy, there's a description of each one at the bottom. And you can see how well it's working. So in this case, the fast algorithm is finding too many points. The high-res is kind of okay if that's what you're looking for. Probably be aiming at the brain, but then you need to define your region of interest to be the brain. For example, this algorithm is not doing that. You can just see the difference it makes on different images. So this is an example of a ischemic stroke. So again, the fast, still processing is probably doing too many data points. It's not actually producing anything at all. The high-res found all these beads there, some over here. I don't know why. Yeah, the sheet must be indicating more points. So yeah, so you have to obviously tune these things up to the other side that you have. There's a bunch of blocks as well. You have some silly images being generated by a dally three. Some of them are not too bad. Some pretty horrendous. Surprising you'll get some relevant ads. I don't know about that. It's not sponsored by it. It's just Google ads. I think I covered this one before. We have a standard, not standard. This is from an implant EEG 16 channels. You can scroll through the data and see what the frequency spectrum looks like. And we have a similar tool that actually convert the EEG into music. So I don't know if it sounds like music to you, right? This one. Give me an ad to learn how to play piano in only 21 days. Again, not sponsored by that company, whatever it is, school, music school. You might also need some more development, but essentially the auto volume will make the volume very quiet when there is no seizure. I cut the end of this data sample. And we know there is a seizure in the middle, so it will become louder. The auto duration will adjust itself to the window size. By the way, yeah, we need to record. So I do record the instruction video of how I'm using this tool and pop it on the website at the bottom. So yeah, because no one reads text. Well, at least I don't assume others don't read anything as well. So it will be good to have a video showing how to use it. But we might be doing more work on it, so it will change anyhow. Yeah, there's a description of what the project is all about by GPT-4. Yeah, the podmaster found the about section on the site. And yeah, it does mention that there is not currently an option to upload data. If you do want me to look at any data or for any of your data to be featured on the site, please go upload it onto Github or any other public repository. We're also looking at the another thing we were thinking adding a couple of a couple of extra topics to the mix. So currently we're doing the bot that plays the cardio game. It was have was generally looking will be trying to open review of a public biomedical data set. Like currently this data on the page is actually using data from IEG.org. I think anyone can after a simple registration can go and get this data. So as I said, we'll be looking at the different biomedical datasets, like the IEG, but the other ones as well. If you haven't tried by chaos.com yet, please go check out the site. Don't forget to provide your feedback.