Neural Mechanism Underlying Orientation Average Representation

When presented with a set of similar stimuli,our visual system is able to quickly and accurately extract the overall statistical properties of the collection(rather than the individual stimulus properties)to form a statistical summary representation.The average value is an important indicator of statistical summary representation.For the neural mechanism of average representation,many studies initially speculated through behavioral evidence.In recent years,researchers began to use neuroimaging and neural signal decoding techniques to try to reveal the neural mechanism of average representation.Only a small number of studies have used high-resolution EEG techniques,among which event related potential(ERP)studies only reveal that average representation is an automated process that does not require attention,without paying attention to the entire time process of average representation occurrence,especially the early component C1.At the same time,many studies on neural mechanisms(including magnetoencephalography(MEG)research)are based on set encoding,which compares the differences in neural activity between sets and individual stimuli to obtain statistical summary representations of neural mechanisms.The differences obtained through subtraction may include the neural mechanisms represented by statistical summary representations,but at the same time,the cumulative effects of neural activity caused by differences in the number of sets and individual stimuli cannot be ruled out.Therefore,the current research results on the neural mechanism of average representation are not rich,and there are many important issues that have not been solved,such as what is the time process of average representation processing? Will it trigger the C1 component originating from the primary visual cortex(V1)? The significance of answering these questions lies in:(1)helping to resolve the controversy over the timing of average representation occurrence in behavioral research;(2)Helps to understand the function of the primary visual cortex in the process of visual integration.To reveal the time course of average representation occurrence,this study used high-resolution EEG technology.Due to the good responsiveness of the primary visual cortex to sinusoidal gratings,the selection of grating stimuli in this study helps to observe early visual related neural activity.In addition,the study adopts a visual adaptation paradigm,comparing adaptive and non-adaptive to ensure that the neural processes represented on average are examined under the condition of matching the number of stimuli.For example,if the subject adapts to a set of stimuli composed of multiple gratings with different orientations(such as a set of gratings with an average orientation of 35 °),and the subject can extract the average representation,then when the test phase presents the average orientation of the corresponding set(such as a 35 ° grating),neural activity will experience repetitive inhibition due to the repetition of psychological representations,which is reflected in a decrease in amplitude on various ERP components.Unlike other ERP components,C1 component is considered bottom-up(feedforward),difficult to regulate,and has a clear origin,namely the primary visual cortex(V1).The stimulation presented in the lower field of vision induces a positive C1 wave,while the stimulation presented in the upper field induces a negative C1 wave.Due to the small amplitude of the C1 component,in order to ensure a sufficiently high signal-to-noise ratio,experiments usually require sufficient repetition and stacking times.Therefore,many previous studies have only selected one field of view for examination.In summary,this study is divided into two experiments to explore the neural mechanisms of average representation processing.Experiment 1 presented the stimulus at the lower field of view position,while Experiment 2 presented the stimulus at the upper field of view position.The independent variables are stimulus type(set,individual)and adaptation type(orientation repetition,orientation control),and the dependent variables are neural activity induced during the adaptation and testing stages.The experiment includes two stages:adaptation and testing.The adaptation stage randomly presents two types of stimuli: a set or a single stimulus,while the testing stage presents a control orientation that is consistent with adaptation(the average orientation of the set or the same orientation as a single stimulus)or a control orientation that is 90 ° from adaptation.The subject’s task is to detect changes in the central fixation point to ensure that attention is always focused on the central fixation point.The analysis results showed that in the time domain,during the testing phase,both experiments found orientation adaptation effects in the early occipital lobe(C1 or P1)and the late frontoparietal lobe(P3 or N2)stages,that is,the amplitude induced by orientation repetition conditions was significantly smaller than that of orientation control conditions.In the adaptation stage,both experiments found that collective stimuli had stronger neural activity than individual stimuli,indicating that the neural activity encoded by the set contained cumulative signals of neural activity caused by the number of stimuli.On the spectrum,the results of the two experiments in the adaptation and testing stages are similar.For θ In the frequency band,the neural oscillation of collective stimuli is stronger than that of individual stimuli;For α In the frequency band,the oscillation of a single stimulus is stronger than that of a collective stimulus.At the same time,both experiments also showed significant time-frequency analysis results during the adaptation stage.For the ERP-related time-frequency signal θ The frequency band(150-250 ms in experiment 1 and80-250 ms in experiment 2)has a Main effect of the stimulus type,and the collective stimulus is stronger than the single stimulus;For ERD signals,in the α In the frequency band(250-400ms),both experiments also found that the Main effect of the stimulus type was significant,and the collective stimulus was stronger than the single stimulus.In summary,our research reveals that the processing of average representation may involve both early and late psychological processes,providing an integrated approach for previous controversies about time processes;The average representation may initially trigger activity in the primary visual cortex;Ensemble encoding triggers low-frequency neural oscillatory activity,which may reflect top-down feedback while the brain integrates information.