Feedback to V1: A reverse hierarchy in vision

Chi Hung Juan, Vincent Walsh

Research output: Contribution to journalArticlepeer-review

104 Scopus citations

Abstract

We examined the involvement of visual area V1 in visual detection to assess the role of feedback connections to V1 as proposed in reverse hierarchy theory (Ahissar and Hochstein 2000). In Experiment 1, signal detection was decreased in feature and conjunction detection tasks by repetitive pulse TMS (rTMS) over V1 for 500 ms after stimulus onset. In Experiment 2, rTMS was delayed to allow uninterrupted signal processing for 100 ms after visual stimulus onset. TMS for the subsequent 500 ms now disrupted detection of conjunction but not feature targets. In Experiment 3 we applied double pulse TMS at varying intervals to assess the timing of V1 involvement in these tasks. Single feature detection involved V1 only at some point between 40 and 100 ms after visual array onset; detection of targets defined by conjunctions of features involved V1 throughout the first 100 ms and also between 200 and 240 ms after visual stimulus onset. We suggest that the early effects in the conjunction task are due to repeated sampling of the visual array to extract the signal from external noise. The later effects in conjunction search are attributed to the return projections from secondary visual areas back to V1, consistent with the reverse hierarchy theory. The effects in both tasks are consistent with early and repeated iterations of feed forward and feedback loops as hypothesised in recent neurophysiological experiments (see Foxe and Simpson 2002).

Original languageEnglish
Pages (from-to)259-263
Number of pages5
JournalExperimental Brain Research
Volume150
Issue number2
DOIs
StatePublished - May 2003

Keywords

  • Delayed onset rTMS
  • Double pulse TMS
  • Reverse hierarchy theory
  • Transcranial magnetic stimulation
  • V1
  • Visual search

Fingerprint

Dive into the research topics of 'Feedback to V1: A reverse hierarchy in vision'. Together they form a unique fingerprint.

Cite this