Related Experiment Videos
Quadrature subunits in directionally selective simple cells: spatiotemporal interactions
1Department of Ophthalmology, University of Rochester, New York 14642, USA.
Visual Neuroscience
|March 1, 1997
Summary
Linear mechanisms alone do not fully explain directional selectivity (DS) in simple cells. A two-subunit model, not a single-branch model, accurately predicts cell responses and nonlinear interactions for DS.
Area of Science:
- Neuroscience
- Computational Neuroscience
- Visual System Physiology
Background:
- Directional selectivity (DS) is crucial for visual processing.
- Linear models of DS in simple cells are gaining popularity.
- Previous models often assume simple linear mechanisms.
Purpose of the Study:
- To investigate if linear mechanisms can explain DS in cat striate cortex simple cells.
- To compare the predictive power of linear models versus a multi-subunit model.
- To determine the necessary complexity for DS in simple cells.
Main Methods:
- Utilized a simple cell with a space-time inseparable receptive field.
- Tested responses to moving bars and nonlinear interactions between pairs of bars.
- Compared experimental data to predictions from 1-branch linear models and a 2-subunit quadrature model.
Main Results:
- Measured cell responses were less modulated and spatially extended than predicted by 1-branch linear models.
- The cell exhibited stronger DS than predicted by linear models.
- Nonlinear interactions and movement responses matched a 2-subunit model, not 1-branch models.
Conclusions:
- Linear mechanisms are insufficient to fully account for DS in this simple cortical cell.
- Multiple nonlinear subunits, specifically a 2-subunit quadrature model, are necessary to explain observed responses.
- The complexity of DS mechanisms in simple cells may be underestimated by simpler linear models.