# Neural Mechanisms for Feature-Based Attention

> **NIH NIH R01** · MASSACHUSETTS INSTITUTE OF TECHNOLOGY · 2024 · $555,564

## Abstract

PROJECT SUMMARY. To find an object in a complex scene, we use feature-based attention to guide our
search, typically in conjunction with spatial attention and targeting eye movements. When searching for our keys
on a table, for example, the features of the keys are used as an attentional template that guides the eyes to the
various objects sharing features with keys until the keys are found. Work from our own and other labs has found
that objects with attended features or attended locations are processed more efficiently in visual cortex, while
the processing of unattended, distracting objects is suppressed. To design an effective neural prosthesis or to
treat people with sensory or attentional impairments, we need a better mechanistic understanding of these
attentional mechanisms at the systems level. The interconnected structures important for the control of attention
have many common features. At the surface level, these common features suggest there may be little difference
in their functions. However, our results show surprising specificity instead. We find that the ventral pre-arcuate
area (VPA) and the frontal eye fields (FEF) in prefrontal cortex (PFC) have different functions in visual search.
Specifically, VPA appears to mediate the selection of likely targets based on their features, and FEF directs
spatial attention and gaze to those possible targets until the object of the search, the target, is found. We
hypothesize that VPA and FEF work together as an interconnected system for guiding gaze to objects we are
searching for, and that they also provide attentional feedback to the occipital and temporal cortex, including the
mid-superior temporal sulcus (mid-STS) region. This feedback biases visual processing in favor of attended
target features. In Aim 1, we will test whether feedback from FEF to visual cortex and VPA is specific for attended
locations, as we propose, or whether the feedback conveys priority values computed from both attended features
and locations. In Aim 2, we will use electrical stimulation coupled with functional magnetic resonance imaging
(fMRI) to finely map the connectome of the mid-STS region, which receives projections from VPA and has been
recently proposed as an important component of the system for the top-down control of attention. The published
connectome will provide a test of anatomical mapping principles that we recently discovered in PFC, and it will
be a valuable resource for the neuroscience community. In Aim 3, we will use neurophysiological recordings in
VPA, FEF, the mid-STS, and other structures proposed to be important for attention, coupled with causal
methods such as optogenetics and muscimol injections to test our hypotheses about the roles of the different
components. One of the key innovations of this project is that we will map the connectivity of the recorded sites
with electrical stimulation and fMRI so that we can target our recordings to the specific neuronal groups in
different struc...

## Key facts

- **NIH application ID:** 10793968
- **Project number:** 2R01EY029666-06
- **Recipient organization:** MASSACHUSETTS INSTITUTE OF TECHNOLOGY
- **Principal Investigator:** Robert Desimone
- **Activity code:** R01 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2024
- **Award amount:** $555,564
- **Award type:** 2
- **Project period:** 2019-01-01 → 2028-12-31

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10793968

## Citation

> US National Institutes of Health, RePORTER application 10793968, Neural Mechanisms for Feature-Based Attention (2R01EY029666-06). Retrieved via AI Analytics 2026-05-24 from https://api.ai-analytics.org/grant/nih/10793968. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*