Detecting synchrony in neuron populations is an important task – often from the perspective of epilepsy and its origins. Several other neurological disorders are linked to hyper- and hyposynchrony, as well. Mostly synchrony is assessed over a whole population of neurons. The z statistic is frequently used to quantify synchrony, especially when thinking of neurons as simple phase oscillators. But – this statistic fails when synchrony is localized to subparts of a population. As in the picture that goes with this post.
So, how would you extract the apparent localized synchrony from data like these? Have a look at the recording of ‘Functional Clustering’, which was part of a presentation for the CAMBAM 2014 Computational Neuroscience Workshop. The presentation was given by Greg Stacey and myself, If you want to know more about the model background, check out Greg’s part, too. Great presentation (as always)!
There are other interesting presentations on the same channel. slowly but surely we are building a great toolbox for analysis here…from students for students.