Lecture Notes on Vision - Third Lecture: Modality of Vision

Phototransduction: Conversion of photons into a neural code.
Retinal Processing: Exploration of the complexity in the retina and subsequent transfer of information to the Lateral Geniculate Nucleus (LGN).
Current Focus: Primary Visual Cortex (V1) and its downstream processing.

Location: V1 is situated in the occipital lobe and is critical for visual functions.
Significance: While V1 is crucial, it is only one of multiple brain areas that respond to visual stimuli. Vision encompasses a vast array of brain components.
Role of V1: V1 follows the LGN in the visual processing pathway and is linked to two main pathways:
- Dorsal Pathway (Where pathway): Concerns spatial location of objects, associated with the magnocellular pathway which handles motion detection.
- Ventral Pathway (What pathway): Relates to object recognition by analyzing color, patterns, orientations, and shapes.

Complexity of Division: Although the delineation between the pathways is useful, it is important to note that interactions (feedback and feedforward) between brain regions blur these lines.
Pathological Evidence: Lesion studies in monkeys and stroke patients provide insights into the distinction between the functional aspects of the two pathways; for instance:
- Lesions in the dorsal pathway may impair spatial location processing.
- Lesions in the ventral pathway may disrupt object identification.

Composition: V1 consists of six layers.
Layer 4: Receives synaptic input from the LGN, with distinct subregions for magnocellular (motion) and parvocellular (fine detail) pathways.
Ocular Dominance Columns: Within Layer 4, there are columns that represent input from either the contralateral or ipsilateral eye, important for segregating visual field data from both eyes.
Cardiocellular Pathways: They bypass Layer 4, synapsing instead in layers 2 and 3 which is significant for complex visual processing.

Types of Cells: Primarily pyramidal and stellate cells facilitate communication within the cortex.
Intrinsic Circuits: Interactions within V1 include feedback loops both from surrounding layers and other brain regions, reflecting the complexity of the visual system.

Stripe Patterning: Imaging studies show distinct ocular dominance in V1, with variations in neuronal response when stimulating one eye versus the other.
Contrast with Other Areas: No similar ocular dominance patterns in V2.

Receptive Fields: Neurons in V1 exhibit selectivity for stimulus orientation; this selectivity can be illustrated through tuning functions, detailing how neuron response varies with orientation.
Tuning Curves: The maximal neuronal response aligns with specific vertical orientations and diminishes for orientations that are further away from the preferred angle.
Differences from LGN Responses: Unlike LGN, V1 neurons respond comparably based on orientation rather than mere presence in the stimulus field.

Modeling Receptive Fields: By combining adjacent center-surround mechanisms, orientation selective fields can be created, permitting the response to oriented stimuli at specific locations in the visual field.
Position Variance: Selectivity captures not only orientations but is position-dependent, where responses vary based on the precise location of stimulus presentation.

Simple Cells: Neurons respond to specific orientations within their receptive field location.
Complex Cells: Exhibit orientation selectivity without position variance; can respond to oriented bars anywhere in their receptive field.
Direction Selectivity: A subcategory where some complex cells respond preferentially to movement in a specific direction, emphasizing the motion aspect of visual processing.

Higher Order Processing: Information integration and contour formation begin in V1 as neurons group together to enhance perceptual salience of visual stimuli.
Contour Saliency Factors: Influenced by the number of coherent line elements, spacing of lines, and overall smoothness of contours. More continuous stimuli generate stronger neuronal responses.

End Stop Cells: Respond to stimuli with inhibitory areas, crucial for detecting line lengths, corners, and turns in visual objects.
Functional Analogy: Studies on dragonfly neurons exemplify how these neurons effectively detect small moving targets (akin to end stop cells) against complex backgrounds.

Models of Direction Selectivity: Differences in processing delays across a network of neurons provides a mechanism for determining moving stimulus direction based on neuronal response timing.
Disparity Selectivity: Neurons can represent depth information through processing the disparity of visual input from both eyes, crucial for three-dimensional perception.

Critical Periods: Neuronal plasticity can be modeled through experimentation in young animals, highlighting the ability of their visual systems to adapt to environments.
Experience-Dependent Plasticity: Training can enhance visual neuron responses, showing adaptability even beyond early development, demonstrating longevity of neural plasticity.

Vector Averaging: A method to deconfound orientation data from various responsive neurons across the visual space to create a coherent perception of orientation that matches stimuli.

Complexity of Visual Neuroscience: This lecture showcases the intricate systems in vision processing, underscoring the need for continual exploration.
Textbook Recommendations: Suggested readings include Candel's textbook, which covers advanced visual processing and object recognition, important for understanding visual neuroscience.