Power-spectra and cross-frequency coupling changes in visual and Audio-visual acquired equivalence learning

Abstract

The three phases of the applied acquired equivalence learning test, i.e. acquisition, retrieval and generalization, investigate the capabilities of humans in associative learning, working memory load and rule-transfer, respectively. Earlier findings denoted the role of different subcortical structures and cortical regions in the visual test. However, there is a lack of information about how multimodal cues modify the EEG-patterns during acquired equivalence learning. To test this we have recorded EEG from 18 healthy volunteers and analyzed the power spectra and the strength of cross-frequency coupling, comparing a unimodal visual-guided and a bimodal, audio-visual-guided paradigm. We found that the changes in the power of the different frequency band oscillations were more critical during the visual paradigm and they showed less synchronized activation compared to the audio-visual paradigm. These findings indicate that multimodal cues require less prominent, but more synchronized cortical contribution, which might be a possible biomarker of forming multimodal associations.

Predicting Stimulus Modality and Working Memory Load During Visual- and Audiovisual-Acquired Equivalence Learning

“…Thus, in the present study, we have focused on what EEG-parameters are important predicting WM-load and/or stimulus modality. How these parameters differ between conditions exceeds the scope of the current study, and has been answered elsewhere (Puszta et al, 2019 ).…”

“…In the current study, we hypothesized that the accuracy of predicting stimulus modality will be high, at least in terms of prediction based on the power and CFC results, as indicated in the previous publication (Puszta et al, 2019 ). Instead of using classical statistical methods, the current study applied machine learning classification algorithms to reveal the EEG features considered important for stimulus modality and/or WM load.…”

“…The EEG datasets previously published by the researchers of the current study and recorded from 18 healthy volunteers during visual and audiovisual associative learning tasks were used (Puszta et al, 2019 ). The task design was established such that associations were learned using the staircase method, thereby enabling the researchers to investigate the number of associations that participants were required to maintain (WM load) and stimulus modality within the same paradigm.…”

“…We excluded five participants due to low signal-to-noise ratios. The current study reanalyzed the EEG data of 18 healthy young adults that were previously published (Puszta et al, 2019 ). The number of samples was identical in both the visual and the audio-visual test.…”

“…The task design was similar to the previous study [for a detailed description, see Puszta and colleagues (Thompson and Paivio, 1994 ; Quak et al, 2015 ) and Eördegh and colleagues (Eördegh et al, 2019 ; Puszta et al, 2018 , 2019 )]. Briefly, the visual and audiovisual tests comprised of three phases—acquisition, retrieval, and generalization.…”