I have two arrays, one representing an auditory stimulus and the other representing neural activity from auditory cortex. The auditory stimulus has known temporal and spectral structure. The neural data has unknown structure, but visual inspection suggests that some stimulus types elicit neural responses with temporal structure similar to that of the stimulus. I'm wondering how best to compare stimulus structure to the response structure in a way that allows me to quantify the degree of structural coupling, thus which stimulus type causes the strongest coupling. I'm assuming that a cross-correlation method could be appropriate, but I'm not sure if it's the best method, or if there are good alternatives.
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
A common method for this is called "reverse correlation", see for example https://scholar.google.com/scholar?hl=en&as_sdt=0,50&q=reverse%20correlation%20receptive%20field
The technique is most popular with visual stimuli (movies, spatiotemporal receptive fields) but can be applied to auditory stimuli (spectrotemporal receptive fields, unfortunately the same "STRF" acronym) as well.
answered Nov 1, 2023 at 0:37