central visual system II

Central Processing of Vision

  • Overview of Visual System

    • Edge Detection

    • The visual system is exquisitely tuned to detect edges, which are fundamental for visual perception. Edges represent critical discontinuities in luminance, color, or texture, forming the fundamental boundaries that define objects in our visual field. They are essential for segmenting scenes, discerning object shapes, extracting depth cues, and guiding our interactions with the environment (e.g., identifying the precise contours of a coffee cup or table for accurate grasping).

    • Notable contributors to this field include David Hubel and Torsten Wiesel, Nobel Prize winners (1981) for their groundbreaking research on the visual system's organization and functionality.

  • Hubel and Wiesel’s Discoveries

    • Discovered specialized neurons in the visual cortex (V1) that are selectively tuned to edges or bars of light of specific orientations and positions within their receptive fields.

    • The significance of edges: Their presence provides critical information for navigating, recognizing objects, and interacting effectively with the environment, as static uniform surfaces convey little information compared to their boundaries.

    • Accidental Discovery:

    • Their pivotal 'accidental discovery' occurred when they struggled to elicit reliable responses from V1 neurons using traditional stimuli like small spots of light or dark projected onto a screen. An unexpected artifact—the sharp shadow of a glass slide being inserted into their stimulus projector—transiently caused vigorous firing in a neuron. This observation revealed that the neuron was responding not to the static spots, but rather to the moving, oriented edge created by the slide. This led them to systematically test oriented lines and edges as stimuli, uncovering the orientation selectivity of cortical neurons.

    • Method: An electrode was used to detect neuronal electrical activity; audible 'pops' through a loudspeaker provided real-time feedback, indicating action potentials of the firing neuron. This allowed immediate identification of effective stimuli.

    • The experimental method highlighted:

      • Initial attempts involved projecting stationary or moving light spots onto the retina, which often produced weak or no responses in certain V1 neurons.

      • The essential finding was that neurons responded robustly and selectively to moving edges or bars of light, particularly those caused by the slight misalignment or sliding of the glass slide across the projector aperture, thereby highlighting the importance of oriented stimuli over isolated spots.

Neuron Response Characteristics

  • Simplified Diagram of Visual Neurons

    • Visual stimuli, such as lines at various orientations, induce varying responses in visual cortex neurons, demonstrating their selectivity.

    • Tuning Curve:

    • A graphical representation illustrating how a neuron's firing rate (response strength) varies in relation to different stimulus orientations. This curve typically shows maximal firing at a particular 'preferred' orientation (e.g., a robust and rapid spike rate at zero degrees, meaning a horizontal bar), with responses gradually decreasing as the stimulus orientation deviates from this optimum.

  • Neuronal Mechanisms and Transitions

    • Neurons in the retina (ganglion cells) and thalamus (Lateral Geniculate Nucleus, LGN) possess center-surround receptive fields, meaning they respond maximally to light in the center and inhibition in the surround, or vice-versa, which is effective for detecting contrast.

    • The transition from these center-surround receptive fields to the tuned edges observed in V1 neurons is a fundamental, albeit complex, transformation. While extensively studied, the precise cellular and synaptic mechanisms are still areas of active research.

    • Proposed mechanism: A widely accepted model for this transition suggests that a single V1 simple cell receives convergent excitatory input from several LGN neurons whose center-surround receptive fields are precisely aligned along a particular axis in visual space. For instance, three ON-center LGN neurons, positioned sequentially, could collectively activate a V1 simple cell most effectively when an oriented bar of light passes through all their excitatory centers, thereby creating an elongated excitatory region characteristic of an orientation-tuned V1 neuron.

Orientation and Spatial Selectivity

  • Layered Structure of Visual Cortex

    • The primary visual cortex (V1) is organized into functional columns, where neurons vertically stacked within a column tend to share similar response properties, such as preferred orientation.

    • These neuronal connections within a column reinforce and sharpen their orientation tuning, creating a highly specific response property.

    • As electrode recordings systematically progress horizontally across the cortical surface, a systematic and orderly change in the preferred orientation of neurons is observed, demonstrating a continuous mapping of orientation sensitivities across the cortex. This means adjacent columns tend to prefer subtly different orientations.

    • Columns also exhibit systematic variations in their response depending on specific spatially localized visual input, contributing to local processing.

  • Important Neurophysiological Concepts

    • Simple Cells: These are V1 neurons characterized by distinct excitatory and inhibitory subregions within their receptive fields, making them highly sensitive to specific orientations and locations of light or dark bars. They respond best to a bar of light (or dark) presented at a particular angle and specific position.

    • E.g., A simple cell may spike vigorously to a horizontally oriented line positioned precisely in the center of its receptive field but will show little or no response if the line is moved slightly up, down, or if its orientation shifts dramatically. This spatial specificity arises from their elongated excitatory and inhibitory subregions.

    • Complex Cells: These V1 neurons also respond to specific orientations but are largely insensitive to the exact position of the stimulus within their broader receptive fields. Unlike simple cells, they lack distinct ON and OFF subregions.

    • For instance, a complex cell may spike robustly to a horizontal line moving anywhere within its receptive field, demonstrating a capacity for generalization across spatial locations. This property is thought to arise from receiving convergent input from multiple simple cells that share similar orientation tuning but have slightly offset receptive field positions.

Columnar and Retinotopic Organization

  • Retinotopy in the visual cortex: This refers to the systematic mapping of the visual field onto the surface of the visual cortex. Adjacent points in the visual space are processed by adjacent neurons in V1, preserving the spatial layout of the visual world.

  • Pinwheel Organization: Within V1, patterns of orientation tuning preferences are organized in a stunning 'pinwheel' fashion. Orientation selectivity radiates systematically from central points (called singularities or pinwheel centers), with all possible orientation preferences represented in a circular fashion around these points.

  • Ocular Dominance Columns: These are alternating stripes of neurons in V1 that preferentially respond to input from one eye or the other, rather than both equally. This columnar connectivity is established from the LGN, with layer IV of the visual cortex exhibiting predominantly monocular response properties—distinctly responding to either the left or right visual field input.

  • Adjacent to the orientation selective columns and ocular dominance columns are blobs. Blobs are metabolically active regions sparsely distributed throughout V1, which are particularly responsive to color and brightness information, but are largely insensitive to orientation. They demonstrate further specialization of visual functions within the primary visual cortex, contributing to various aspects of early visual processing beyond just form and motion.

Visual Pathways and Their Functions

  • Parallel Processing Streams: Beyond V1, visual information diverges into two primary processing streams, allowing for parallel, specialized analysis of different visual features:

    • Dorsal Stream: Also known as the "Where/How" pathway, this stream projects from V1 through the parietal cortex. It is crucially important for processing spatial information (locating objects in space), motion perception, judging depth, and guiding visually-motor actions (e.g., reaching, grasping, and navigating). Dysfunction here leads to difficulties in spatial awareness and action execution despite intact object recognition.

    • Ventral Stream: Often referred to as the "What" pathway, this stream travels from V1 through the temporal cortex. It is essential for object recognition, processing visual features such as shape, texture, and color. It allows us to identify what objects are, including categorizing them (e.g., recognizing a face as a specific person, or a chair as furniture).

  • Clinical Examples

    • Posterior Cortical Atrophy (PCA):

    • A neurodegenerative condition, often considered a rare variant of Alzheimer's disease, that predominantly affects the dorsal stream by causing atrophy in the posterior parietal and occipital cortices. This leads to profound difficulties in spatial localization, visual-motor coordination, and navigation, despite relatively preserved visual acuity and object recognition in early stages.

    • Patient example: An individual with PCA might exhibit observable inability to accurately touch or grasp an object upon request (ataxia), or struggle to judge distances, despite being fully able to identify and describe the object visually. Navigation and dressing difficulties are also common.

    • Prosopagnosia:

    • A neurological disorder primarily affecting the ventral stream, specifically the fusiform gyrus (often associated with the Fusiform Face Area, FFA), leading to a severe impairment in face recognition. This condition can be congenital (developmental) or acquired due to brain injury.

    • Example: A patient with prosopagnosia might struggle immensely to recognize even close family members or their own face from visual input alone but can readily identify them by non-facial cues such as their voice, distinctive clothing, or gait. They can see and describe facial features but cannot link them together to form a coherent identity.

Conclusion: Integration of Visual Processing

  • The visual system employs a sophisticated strategy of 'divide and conquer,' processing diverse visual attributes such as form, motion, orientation, and color in specialized, parallel neural circuitry. This modularity ensures efficient and detailed analysis of complex visual scenes.

  • Insights from pioneers like Hubel and Wiesel’s work, combined with contemporary understanding of cortical architecture, illustrate how these distinct modules within the visual cortex collaborate and integrate their processed information. This intricate integration is fundamental to constructing a coherent, comprehensible, and unified perception of the visual world, enabling us to interact effectively with our environment.