New paper in press in elife: Waschke et al.

Oble­ser­lab senior PhD stu­dent Leo Waschke, along­side co-authors Sarah Tune and Jonas Obleser, has a new paper in eLife.

The pro­cess­ing of sen­so­ry infor­ma­tion from our envi­ron­ment is not con­stant but rather varies with changes in ongo­ing brain activ­i­ty, or brain states. Thus, also the acu­ity of per­cep­tu­al deci­sions depends on the brain state dur­ing which sen­so­ry infor­ma­tion is processed. Recent work in non-human ani­mals sug­gests two key process­es that shape brain states rel­e­vant for sen­so­ry pro­cess­ing and per­cep­tu­al per­for­mance. On the one hand, the momen­tary lev­el of neur­al desyn­chro­niza­tion in sen­so­ry cor­ti­cal areas has been shown to impact neur­al rep­re­sen­ta­tions of sen­so­ry input and relat­ed per­for­mance. On the oth­er hand, the cur­rent lev­el of arousal and relat­ed nora­dren­er­gic activ­i­ty has been linked to changes in sen­so­ry pro­cess­ing and per­cep­tu­al acu­ity.

How­ev­er, it is unclear at present, whether local neur­al desyn­chro­niza­tion and arousal pose dis­tinct brain states that entail vary­ing con­se­quences for sen­so­ry pro­cess­ing and behav­iour or if they rep­re­sent two inter­re­lat­ed man­i­fes­ta­tions of ongo­ing brain activ­i­ty and joint­ly affect behav­iour. Fur­ther­more, the exact shape of the rela­tion­ship between per­cep­tu­al per­for­mance and each of both brain states mark­ers (e.g. lin­ear vs. qua­drat­ic) is unclear at present.

In order to trans­fer find­ings from ani­mal phys­i­ol­o­gy to human cog­ni­tive neu­ro­science and test the exact shape of unique as well as shared influ­ences of local cor­ti­cal desyn­chro­niza­tion and glob­al arousal on sen­so­ry pro­cess­ing and per­cep­tu­al per­for­mance, we record­ed elec­troen­cephalog­ra­phy and pupil­lom­e­try in 25 human par­tic­i­pants while they per­formed a chal­leng­ing audi­to­ry dis­crim­i­na­tion task.

Impor­tant­ly, audi­to­ry stim­uli were selec­tive­ly pre­sent­ed dur­ing peri­ods of espe­cial­ly high or low audi­to­ry cor­ti­cal desyn­chro­niza­tion as approx­i­mat­ed by an infor­ma­tion the­o­ret­ic mea­sure of time-series com­plex­i­ty (weight­ed per­mu­ta­tion entropy). By means of a closed-loop real time set­up we were not only able to present stim­uli dur­ing dif­fer­ent desyn­chro­niza­tion states but also made sure to sam­ple the whole dis­tri­b­u­tion of such states, a pre­req­ui­site for the accu­rate assess­ment of brain-behav­iour rela­tion­ships. The record­ed pupil­lom­e­try data addi­tion­al­ly enabled us to draw infer­ences regard­ing the cur­rent lev­el of arousal due to the estab­lished link between nora­dren­er­gic activ­i­ty and pupil size.

 

Sin­gle tri­al analy­ses of EEG activ­i­ty, pupil­lom­e­try and behav­iour revealed clear­ly dis­so­cia­ble influ­ences of both brain state mark­ers on ongo­ing brain activ­i­ty, ear­ly sound-relat­ed activ­i­ty and behav­iour. High desyn­chro­niza­tion states were char­ac­ter­ized by a pro­nounced reduc­tion in oscil­la­to­ry pow­er across a wide fre­quen­cy range while high arousal states coin­cid­ed with a decrease in oscil­la­to­ry pow­er that was lim­it­ed to high fre­quen­cies. Sim­i­lar­ly, ear­ly sound-evoked activ­i­ty was dif­fer­en­tial­ly impact­ed by audi­to­ry cor­ti­cal desyn­chro­niza­tion and pupil-linked arousal. Phase-locked respons­es and evoked gam­ma pow­er increased with local desyn­chro­niza­tion with a ten­den­cy to sat­u­rate at inter­me­di­ate lev­els. Post-stim­u­lus low fre­quen­cy pow­er on the oth­er hand, increased with pupil-linked arousal.

Most impor­tant­ly, local desyn­chro­niza­tion and pupil-linked arousal dis­played dif­fer­ent rela­tion­ships with per­cep­tu­al per­for­mance. While par­tic­i­pants per­formed fastest and least biased fol­low­ing inter­me­di­ate lev­els of audi­to­ry cor­ti­cal desyn­chro­niza­tion, inter­me­di­ate lev­els of pupil-linked arousal were asso­ci­at­ed with high­est sen­si­tiv­i­ty. Thus, although both process­es pose behav­ioural­ly rel­e­vant brain states that affect per­cep­tu­al per­for­mance fol­low­ing an invert­ed u, they impact dis­tinct sub­do­mains of behav­iour. Tak­en togeth­er, our results speak to a mod­el in which inde­pen­dent states of local desyn­chro­niza­tion and glob­al arousal joint­ly shape states for opti­mal sen­so­ry pro­cess­ing and per­cep­tu­al per­for­mance. The pub­lished man­u­script includ­ing all sup­ple­men­tal infor­ma­tion can be found here.