V1 Flashcards
Hubel and Wiesel’s discovery of receptive fields in the striate cortex (V1):
Recorded neural activity in the primary visual cortex (V1) of cats
Used light stimuli presented on a screen while measuring neuron firing
Key Findings:
Unlike retinal or LGN neurons, V1 neurons were not excited by simple spots of light
Instead, they responded best to edges, bars, and lines with:
-Specific orientations
-Specific positions in the visual field
-Movement in a preferred direction (for some)
✅ Conclusion:
V1 neurons are tuned to orientation, not just light intensity — they act as edge detectors.
Binocularity and ocular dominance
In primary visual cortex (V1), inputs from the two eyes are initially segregated in layer 4C, where:
Magnocellular input (motion, depth) → 4C𝛼
Parvocellular input (color, detail) → 4C𝛽
Each LGN layer projects to V1 with monocular (one-eye) input only
Above and below layer 4C, neurons begin to receive input from both eyes — these are called binocular neurons
Ocular Dominance Columns
In layers outside 4C, neurons are organized into ocular dominance columns:
-Alternating bands of neurons in V1 that prefer left or right eye input
-Each column is more strongly driven by one eye, but many neurons are binocular
Role of binocular neurons?
* Perfectly suited to compare the retinal images of the 2 eyes (depth perception/stereoscopic vision), which show slightly different views of the word –> Helps perceive relative depths of objects in the environment
* Binocularity also increases the sensitivity of the visual system
Orientation selectivity
V1 neurons (especially simple cells) respond best to bars or edges of light at a specific orientation (e.g., vertical, horizontal, diagonal).
V1 neurons are also tuned to:
-Spatial frequency (how wide or narrow the stripes are)
-Gratings (patterns of repeated bars)
-Rectangular ON/OFF regions (not circular like retinal/LGN neurons)
How are these orientation-selective receptive fields formed?
👨🔬 Hubel & Wiesel’s Model:
Multiple LGN neurons, each with circular receptive fields, converge onto a single V1 simple cell.
-These LGN inputs are arranged in a line — their overlapping ON regions form a rectangular excitatory zone, surrounded by inhibition.
-🧩 The result:
The V1 neuron now responds best when a bar of light aligns with the axis of this arrangement.
If the bar is tilted or misaligned, fewer LGN inputs are activated → weaker response
✅ This creates a selective response to orientation.
Directional Motion Sensitivity
Some V1 neurons are direction-selective: they respond strongly to motion in one direction, and weakly or not at all in the opposite or orthogonal directions.
-This helps the brain detect object movement and motion flow.
Directional Tuning Curve (Polar Plot)
Plots firing rate (Hz) as a function of movement direction (360° around a circle).
-The farther from the center, the higher the firing rate.
- A neuron with directional preference will have a peak in the direction it prefers.
Tuning Curve Description:
On a polar plot:
180° (downward) = longest line (highest firing rate)
0° or 360° (upward) = shorter line (lower response)
90° and 270° (horizontal) = no line (no response)
Colour contrast detection
- Cells in the parvocellular pathway provide information about red-green light (lower 4C) (4C𝛽)
- Cells in the koniocellular pathway provide information about blue-yellow light (upper 4C layers in blobs) - Directly to blobs between layers (Blobs are color-sensitive regions within layers 2/3 of V1)
Ultimately, these pathways lead to blobs (have solar sensitive cells) in V1
Layers 3,4,5,6 -> p-ganglion, position/colour sensitive (green, red)
Layers in between m & p -> koniocellular
Simple cells
cortical neurons whose receptive fields have clearly defined inhibitory and excitatory regions
Complex cells
cortical neurons whose receptive fields do not have clearly defined inhibitory and excitatory regions
* E.g., may respond to a stripe presented anywhere in its receptive field (i.e., it is “phase-insensitive”
ex: A complex cell may respond strongly to a vertical bar, whether it’s left, center, or right within its receptive field.
In contrast, a simple cell would only respond if the bar is in just the right position (phase-sensitive).
Ocular dominance columns
Neurons are organized into vertical columns that prefer input from either the left or right eye.
These columns are arranged in an interdigitated pattern (L–R–L–R…).
Each column spans all layers except layer 4C, where inputs are still segregated.
Distance between adjacent ODCs: ~0.5 mm
neurons are arranged in an interdigitated “left-preferring” and “right-preferring” manner
Orientation Columns
Neurons within a column prefer the same orientation of stimulus (e.g., vertical, 45°, horizontal).
Orientation preference changes gradually across the cortex — forming a map of all possible angles.
These columns are organized in a radial (pinwheel) pattern, with the center of the pinwheel being a key landmark.
old Hubel and Wiesel’s Hypercolumn Model (“Ice Cube Model”)
A hypercolumn:
A ~1 mm² block of V1 that contains:
One full set of orientation columns for each eye
So, 2 ocular dominance columns, each with all orientation preferences
It processes all visual features (form, motion, depth, colour) for a small region of the visual field
Each contains info about a very small part of the retina
* 1 mm block of cortex (each “sees” a different portion of the visual field)
* Distance between centre of adjacent ODCs is ~0.5 mm
Colour Sensitivity and Blobs
🔬 Identified using cytochrome oxidase (CO) staining:
Blobs: clusters of CO-positive cells in layers 2/3 → high metabolic activity
Located in the center of ocular dominance columns (~0.5 mm apart)
Function: likely involved in colour processing
Receive parvocellular and koniocellular input → process fine detail and colour
➖ Interblob regions:
Lie between blobs
Process form and orientation (not colour)
Staining the visual cortex for cytochrome oxidase (CO), an enzyme that labels metabolically-active cells, reveals vertical columns in layer 2/3 called blobs (bands of CO positive cells)
Occur in centre of ocular dominance columns (~0.5 mm apart)
Function of blobs may be to process colour information
Blob and interblob regions are anatomically distinct
Linked to visual processing through 2 separate visual systems: magno and parvo (preferentially to blobs, cones: detail/colour)
updated hypercolumn model
Researchers use intrinsic signal imaging to measure brain activation:
Shine constant light on cortex → measure reflected light
Active regions absorb more light (due to deoxygenated hemoglobin), appear darker
Orientation columns are arranged in a pinwheel structure
Colour blobs are located at the centers of these pinwheels
Reinforces idea that orientation and colour are processed in parallel but spatially organized systems
Hypercolumn: a 1 mm block of striate cortex contains “all the machinery necessary to look after everything the visual cortex is responsible for, in a certain small part of the visual world” (Hubel, 1982)
2 ocular dominance columns, ODCs (L, R)
Each ODC covers every possible orientation
The centre of the orientation pinwheel contains a colour-sensitive “blob”
Adaptation and tilt aftereffect
A reduction in neural response after prolonged or repeated exposure to a stimulus. It’s a powerful tool to probe how neurons are tuned to specific stimulus features like orientation.
Before Adaptation (Left Panel)
Each bar = a group of V1 neurons tuned to a specific line orientation (e.g., –10°, 0°, 10°, etc.).
When a vertical line (0°) is shown, neurons tuned to 0° respond most strongly, with nearby orientation-tuned neurons responding less.
After Adaptation (Right Panel)
The subject is exposed to 20° stripes for a prolonged period. Neurons tuned to 20° become fatigued, reducing their responsiveness.
-10° and 0° neurons are also slightly fatigued.
When a vertical (0°) bar is shown again: The –10° neurons (not fatigued) fire relatively more.
This shifts the population response toward the left.
Tilt AfterAffect: The brain interprets the shifted activity pattern as if the stimulus is tilted slightly leftward → this is a tilt aftereffect.
You perceive the vertical line as being tilted left (even though it’s not), because of the adaptation-induced imbalance in neural responses
The tilt aftereffect
- Perceptual illusion of tilt, produced by adaptation to a pattern at a given orientation
- Supports the idea that human visual system also has individual neurons selective for different orientations
- after adaptation to tilted line stimulus, straight lines should appear to point left (if you do this with one eye closed, the tilt effect still occurs
Ganglion Cells and Stripes
Retinal ganglion cells respond best to gratings (striped patterns) with a specific spatial frequency that matches the size of their receptive field.
Low spatial frequency (broad stripes):
Light and dark areas fall across both center and surround → cancels out → weak response
Medium spatial frequency (just right):
Aligns with center-surround structure → strongest response
High spatial frequency (fine stripes):
Too small → averages out over receptive field → weak response
Each cell acts like a “Goldilocks filter” → not too big, not too small, just right
Adaptation to Spatial Frequency experiment
🔬 Procedure:
1.Observe a grating of a specific spatial frequency (e.g., 7 cycles/degree)
2.After 20 seconds, look at a neutral stimulus
- Your contrast sensitivity function (CSF) shifts — you temporarily lose sensitivity at that frequency
What This Shows:
The visual system has neurons tuned to different spatial frequencies
-Adaptation reduces the responsiveness of those neurons → leading to a dip in contrast sensitivity at that frequency
After adapting to 7 cpd: Your brain is less sensitive to patterns at 7 cpd
But still sensitive to higher or lower frequencies
Recordings from monkey striate cortex reveal neurons with different spatial-frequency tuning functions – all neurons work together
Where are the neurons that encode adaptation?
Interocular Transfer Experiment:
You adapt to a stimulus (e.g., a grating) using only one eye (e.g., left eye). Then you test for the adaptation effect using the other eye (e.g., right eye).
What Happens?
Adaptation still occurs! Even though the test eye never saw the adapting stimulus, there’s a reduction in sensitivity (e.g., tilt aftereffect or spatial frequency dip is still perceived).
What Does This Mean?
Adaptation must happen in neurons that receive input from both eyes → i.e., binocular neurons
These are found in the primary visual cortex (V1) and beyond, not earlier in the visual pathway. Adaptation occurs in V1 or later in the visual pathway, where inputs from both eyes converge onto binocular neurons.
interocular transfer
Transfer of adaptation from adapted to non adapted eye is called interocular transfer
- Because inputs from the two eyes don’t converge until V1, adaptation must happen in cortical neurons (V1 or beyond)!
Spatial Frequency–Tuned Pattern Analyzers
What Are They?
The visual system breaks down images using spatial frequency channels. Each channel is a group of cortical neurons tuned to respond best to a specific range of spatial frequencies. This concept is like a neural filter bank — different neurons are sensitive to different levels of visual detail.
Types of Frequency Channels:
1. Low spatial frequency: Broad outlines, general shapes, large-scale contrast
2. High spatial frequency: Fine detail, edges, sharp boundaries
Spatial frequency channel is a pattern analyzer, implemented by an ensemble of cortical neurons, in which each set of neurons is tuned to a limited range of spatial
frequencies
How we study vision in infants
Preferential looking
- Present infants with 2 stimuli
- Measure time spent staring at each stimulus - can the baby still see w/decreased contrast?
- Babies will preferentially look at grading
- Problems with this method? Baby will get distracted, not cooperate based on other factors, how hungry/tired they are
How we study vision in infants
Measure visually-evoked potentials (VEPs)
- Attach electrodes to infant’s scalp and measure changes in electrical activity in the underlying cortex
- Present infants with 2 stimuli
- Measure electrical signals from the brain, very quick.
- start with thick bars (low frequency) and then decrease thickness (higher frequency)
baby Visual development
The rod system is functional in early infancy
Cone system: Infants < 1 month can’t discriminate colours; this ability emerges by 2-3 months
Visual acuity and contrast sensitivity for high spatial frequencies develop slowly and may not reach adult levels until several years of age… babies are not good at detecting detail/contrast (can see difference between black and white)
Development of the CSF - Contrast Sensitivity Function
- Sensitivity to low spatial frequencies develops sooner
- Postnatal changes in the retina limit the development of acuity and contrast sensitivity
- Foveal receptor density increases after birth → finer cone sampling in later childhood (cones develop after birth, density increases and shape, size and packing changes) happens after birth to contribute/inc. visual acuity
- Migration of RGCs and inner nuclear layers away from the foveal region (foveal pit) → fovea isn’t fully adultlike until 4 years of age!
20’s we are most sensitive, loose contrast sensitivity at high frequency as we grow
Consequences of abnormal visual experience during development
Cataract: opacity of the lens (light reaching both eyes is not equal, one becomes more blurry)
Strabismus: one eye is turned so that the two eyes see different views of the world (eyes are not aligned/ not working together)
Anisometropia: two eyes have very different refractive errors (one eye has different refractive error than the other. One can be very near sighted and the other is normal which causes a big disbalance and impaired development)
