Early recurrence and ongoing parietal driving during elementary visual processing

Plomp, Gijs; Hervais-Adelman, Alexis; Astolfi, Laura; Michel, Christoph M. · 2015 · Crossref

DOI: 10.1038/srep18733

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates the timing and directionality of neural interactions during early visual processing, specifically addressing whether recurrent feedback and top-down attentional influences occur at short latencies (<100 ms) after stimulus onset. Traditionally, activity in this window is attributed to bottom-up feed-forward propagation from primary visual cortex (V1). However, the reciprocal anatomical connectivity between visual areas suggests that fast, two-way interactions are possible. The authors aimed to distinguish between these processes in humans by combining high-temporal-resolution EEG source imaging with Granger-causal modeling. The researchers presented task-irrelevant checkerboard stimuli in the lower visual fields while participants performed a central fixation task. They localized twelve regions of interest (ROIs) in occipital, parietal, and frontal cortices for each participant using fMRI. EEG data were then analyzed using weighted Partial Directed Coherence (wPDC) to estimate directed connectivity between these ROIs. This approach allowed the authors to quantify the strength and direction of influence between areas at specific time points, focusing on the C1 component (peaking at 76 ms) and the N1 component (peaking at 146 ms). The results revealed distinct patterns of connectivity. At the C1 latency, V1, MT, and LOC showed increased driving for contralateral stimuli, consistent with feed-forward processing. Crucially, the analysis identified fast, two-way interactions between MT and V1, with feedback from MT to V1 occurring simultaneously with feed-forward signals. This suggests that MT influences V1 activity within the first 100 ms, potentially via indirect pathways bypassing direct V1 input. Additionally, the lateral intraparietal cortex (LIP) exhibited the strongest summed driving across the network at both C1 and N1 latencies. Unlike the visual areas, LIP’s influence was not stimulus-specific and was most prominent in the alpha band, indicating an ongoing, top-down role in coordinating visual activity. These findings challenge the view that early visual processing is purely bottom-up. The study demonstrates that recurrent interactions between higher-level visual areas (MT) and V1 occur rapidly, likely aiding in the disambiguation of visual transients. Furthermore, the dominant and continuous driving from LIP suggests that top-down attentional mechanisms are active from stimulus onset, modulating the entire visual network rather than targeting specific areas. The work establishes that combining EEG source imaging with Granger-causal modeling can effectively separate fast feedback loops from sustained top-down control, providing a refined understanding of the dynamic, interactive nature of human visual processing.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• useful field of view