Comparing Crowding and Foveal Feedback mechanisms using a Vernier task

Feedback connections from higher to lower visual areas have shown to play a role in visual perception by modulating attention, simplifying stimulus perception, and providing more detailed representation of objects in the visual field. An fMRI study by Williams et al. (2008) showed that a representation of peripheral stimuli is available in the foveal retinotopic cortex during a shape discrimination tasks, suggesting a new form of feedback. Following investigations showed that this feedback process can be perturbed: when noise is presented in the foveal visual field within a delay from stimulus onset (~150 ms for some studies, 250 ms for others), performance is impaired. Recent evidence has revealed that the well-known mechanism of crowding also seems to have a critical timing in which target-flanker interactions take place. The strongest crowding effects can be observed when the flankers appear with a temporal delay between 100 and 150 ms after the onset of the target (Tkacz-Domb & Yeshurun, 2017; Yeshurun et al., 2015). This U- shaped crowding function suggests strong similarities with what typically happens in metacontrast masking, where a mask can compromise the coding of a peripheral target when presented in the same spatial location but with a delay of about 50-150 ms (Chung, 2016; Chung & Patel, 2022; Di Lollo et al., 2000; Harrison & Bex, 2014; Huckauf & Heller, 2004). What do these predominantly peripheral visual mechanisms have in common? And how can we study them? In this work, we wanted to test the hypothesis that foveal feedback only occurs for high level objects by replicating the previous studies with low level stimuli and investigate the temporal and spatial properties of this phenomenon. A first experiment adapted the paradigm of previous studies to investigate if the same phenomenon would also apply to lower-level stimuli. In a second experiment we combined standard foveal feedback paradigms with crowding to better investigate their nature and the relationship between the two mechanisms. In the final and third experiment we tested the characteristics of the foveal noise to understand whether the disturbances could be caused by a simple, lower-level foveal distractor or if they are dependent of the noise characteristics. We observed a peak in Vernier acuity threshold between 50 and 150 ms after stimulus onset in the first two experiments, while the third one did not produce significant results. We propose that the effect of a foveal disruptor in a peripheral visual acuity task is task-dependent, and it cannot be induced by the simple presence of any type of delayed foveal object. These findings suggest a potential link to the concept of cortical reentrant processing (Di Lollo et al., 2000), with recurrent and mechanisms playing a pivotal role in object detection and recognition within precise temporal windows.