search

kemerelab / jagular

Out-of-core pre-processing of big-ish electrophysiology data
Other
3 stars 1 forks source link

implement: spike detection #8

Open eackermann opened 7 years ago

eackermann commented 7 years ago

MAD-based spike detection.

We should think carefully about lock-out periods, peeling, and adaptive detection.

For adaptive detection, we could use the block size, and do detection on a per-block level. However, we will again need to have some padding to ensure that we don't cut some spikes in half, and that we don't detect the same spikes multiple times.

eackermann commented 7 years ago

This is also a useful reference / starting point: http://xtof.perso.math.cnrs.fr/locust_sorting_python.html