Thanks, that's interesting. Google gave http://chaos.c.u-tokyo.ac.jp/papers/...os_chaotic.pdf
. Do you think the heteroclinic channel http://www.ploscompbiol.org/article/...l.pcbi.1000072
is a simple (maybe non-chaotic version) of chaotic itinerancy? Actually, now reading the second article more carefully, in the section "Mathenatical Image and Models", they do explicitly discuss chaotic itinerancy as being similar, topologically, but they weren't able to find parameters that made it work for their application.