Thread by @paulmatusz

Best things come in 3. Congrats to @NoraPlethora for her 3rd PhD pub, in @NeuroImage_EiC . We used N2pc as an est'd marker of attn control & tested how visual goals, multisens salience & context (object predictability & meaning) interact in controling capture by distractors 1/n

https://twitter.com/seeingwithsound/status/1437385169074368517

We were interested in the role of lateralised brain mechs in those interactions (lat. brn mechs are 'captured' by canonical N2pc analyses) & we tested them w canonical N2pc & multivariate el.nimg anals. In fact lateralised N2pc-like mechs did NOT support those interactions 2/n

Motivated by our prev work showing the role of NON-lateralised brn mechs in multisensory attention, we tested also for those brain mechs. As both object predictions & semantics affect stim processing fr early on we also extended the anls window from stim onset to that of N2pc 3/n

Finally we set to understand if the brn mechs are strength-based ("gain control" ) and/or topographic, netwrk-bsd. Attention-context interactions modlt'd distr re's from 30ms on,"relaying" meaningful & arbitrary vis & AV distr prcsng to distnct networks, up to N2pc stages 4/n

Strength-bsd brn mechs became active mid-latency, >100ms & later. In behavior we found that attention capture by meaningful distractors was suppressed compared to arbitrary distractors. Thus, meaning can be, next to predictions, 2nd context mech utilised by top-down attn 5/n

This was an ambitious project (props to @NoraPlethora) aiming to clarify how we pay attention in outside world, by 'bringing outside world to the lab' via emulating factors affecting attn outside. It was vital to use N2pc & cuing effs as "starting points" to study & interpret 6/n

intercts of goals,AV salience & context & to use multivar anls to clarify the role of lat. vs non-lat. brain mechs in controling 'naturalistic' attention. This was truly a team effort w @MicahMMurray @tivadarux & Chrysa Retsa, supp'd by @snsf_ch, @MedGIFT_group & @ICT_Sierre 7/n

Pls unroll thread @threadreaderapp