Kilosort VS fooof

And yet, we are starting to make sense of it all! The experimental tools are really what makes the difference. Lots of recently developed genetic tools (such as optogenetics ) are now allowing us for the first time to activate/inactivate specific neurons and molecular pathways and observe the causal effects . Other genetic methods (like Crispr) allow us to create mutant animals where a specific gene is mutated, so a lot of work is now in rodents, but I expect it to translate to humans in the next decade or so. Another really crazy one are cerebral organoids, where we grow in vitro a simplified human brain from stem cells to study the details of development and the function of different genetic pathways. When it comes to brain function, the ability to record from tens of thousands of neurons simultaneously using new probes like Neuropixels has been a recent game changer. Also, machine learning approaches are now incredibly useful to analyze data, from detecting the individual "spikes" to developing complex models of the brain dynamics to support computation. To give you a concrete example of progress, we can now read a person's mind so he can write words on a screen, pretty much as quickly as if he was typing them on a keyboard.

I'm assuming by clustering neurons, you mean clustering spike data. There are several excellent options that already exist for this purpose, the most notable of which is KiloSort.

If you have experience with Python, you can use the FOOOF (Fitting Oscillations & One Over F) toolbox to analyze your EEG data and find the highest peak in the theta band for each electrode.