Cours NSC6070 : Physiologie et Pathologies de la Vision

General Anatomy of the Eye and Retinal Structure

Visual perception begins with light refraction through the cornea and lens, which focuses rays onto the retina, specifically targeting the fovea for maximum resolution. The retina is a complex multi-layered tissue composed of the pigment epithelium, followed by a layer of photoreceptors (rods and cones), and then middle and inner layers containing horizontal, bipolar, amacrine, and ganglion cells. Axons from the ganglion cells aggregate to form the optic nerve. The fovea is characterized by a high packing density of cones, approximately $180,000$ cones, and the exclusion of other cell layers to minimize light scattering, whereas the foveola represents the very center of this region. Rods are absent in the central fovea but reach their highest density in the periphery, roughly $20^\rightleftharpoons$ from the fovea. The optic disc, where the optic nerve exits, lacks photoreceptors entirely, creating a physiological blind spot.

Functional Characteristics of Rods and Cones

There are two primary types of photoreceptors with distinct morphological and physiological properties. Rods contains free-floating discs in their outer segments and are characterized by extreme sensitivity to light, capable of responding to a single photon. This makes them ideal for scotopic (low-light) vision, though they offer very low spatial resolution and do not contribute to color perception. In contrast, cones have outer segments formed by the folding of the cell membrane and require a higher light threshold (>100 photons) to activate. They provide high spatial resolution and mediate color vision through three subtypes: S-cones ($\text{Short-wavelength, blue, peak } \backsim 420-430,nm$), M-cones ($\text{Medium-wavelength, green, peak } \backsim 530,nm$), and L-cones ($\text{Long-wavelength, red, peak } \backsim 560,nm$). Rods have a peak spectral absorbance at roughly $500,nm$.

The Mechanism of Phototransduction

Unlike most sensory receptors that depolarize in response to a stimulus, photoreceptors are depolarized in the dark and hyperpolarize when exposed to light. In darkness, rhodopsin is inactive, and high levels of cytoplasmic cGMP keep sodium ($Na^+$) channels open, allowing a steady influx of ions (the dark current) which maintains the cell at a potential of approximately $-40,mV$. This state promotes the continuous release of the neurotransmitter glutamate. When light strikes, rhodopsin becomes active and triggers a biochemical cascade involving the G-protein transducin, which activates phosphodiesterase (PDE). PDE breaks down cGMP into GMP, causing the $Na^+$ channels to close. The resulting reduction in positive ion influx leads to hyperpolarization (reaching about $-70,mV$), which significantly decreases the release of glutamate. Recovery of the photovoltage typically occurs within approximately $0.2,s$ after a brief light flash.

Retinal Signal Processing: ON and OFF Pathways

Information processing in the retina is based on detecting variations in luminance. Bipolar and ganglion cells are categorized into ON and OFF types based on their response to glutamate. OFF bipolar cells express ionotropic AMPA or kainate receptors; since they are excited by glutamate, they depolarize in the dark and hyperpolarize in the light. ON bipolar cells express metabotropic mGluR6 receptors; glutamate binding to these receptors triggers a cascade that closes $Na^+$ channels. Therefore, in the dark, they are inhibited (hyperpolarized), but when light reduces glutamate release, they are disinhibited and depolarize. This dual-system architecture allows the visual system to detect both light objects on dark backgrounds (ON cells) and dark objects on light backgrounds (OFF cells). Spatial integration is further refined by horizontal cells, which provide lateral inhibition between the central photoreceptor and its neighbors, creating center-surround receptive field organizations.

Anatomy of the Visual Pathways from Retina to Cortex

The visual pathway follows a precise anatomical route: Retina $\rightarrow$ Optic Nerve $\rightarrow$ Optic Chiasm $\rightarrow$ Optic Tract $\rightarrow$ Lateral Geniculate Nucleus (LGN) $\rightarrow$ Optic Radiations $\rightarrow$ Primary Visual Cortex (V1 or Striate Cortex). At the optic chiasm, fibers from the nasal retina (representing the temporal visual field) cross to the contralateral hemisphere, while fibers from the temporal retina remains ipsilateral. The LGN in the thalamus is organized into six distinct layers: layers 1-2 are the Magnocellular (M) layers, receiving input from M-type (Parasol) ganglion cells sensitive to motion and depth; layers 3-6 are the Parvocellular (P) layers, receiving input from P-type (Midget) ganglion cells sensitive to color and fine detail. Koniocellular layers reside ventral to each principal layer. Other retinal projections include the suprachiasmatic nucleus (circadian rhythm regulation), the pretectum (pupillary reflex), and the superior colliculus (eye and head movement orientation).

Functional Organization of the Primary Visual Cortex (V1)

V1 is located along the calcarine sulcus in the occipital lobe and displays a retinotopic organization where the central visual field is heavily over-represented (cortical magnification). It is organized into specialized columns. Ocular dominance columns represent input from either the left or right eye and are established during a postnatal critical period; monocular deprivation during this window can lead to permanent structural changes. V1 also contains orientation-selective neurons. Simple cells respond to bars of light in specific orientations and positions, while complex cells are orientation-selective but position-invariant. Orientation preferences are organized in "pinwheel" configurations. Hypercolumns are functional units (roughly $4,mm$ in size) that contain a complete set of orientation columns and ocular dominance columns for a specific region of space. Additionally, cytochrome oxidase "blobs" within V1 are specialized for color processing.

The Ventral and Dorsal Processing Streams

Higher-level visual processing is divided into two major functional pathways. The Ventral Pathway (the "What" pathway) extends from V1 through V2 and V4 to the inferotemporal cortex (IT). This stream is responsible for object recognition, color, and form analysis. V4 is particularly sensitive to color and fine textures, while IT neurons respond to complex shapes and faces, showing size and position constancy. The Dorsal Pathway (the "Where" or "How" pathway) extends from V1 through V2 and V3A to the middle temporal area (MT/V5) and the posterior parietal cortex. This stream mediates motion perception, spatial localization, and visuomotor coordination. MT neurons have large receptive fields and are highly selective for motion direction, with their activity capable of being organized into directional maps. Current models suggest a hybrid of hierarchical (simple to complex feature integration) and parallel (simultaneous processing of different attributes) processing.

Clinical Deficits and Visual Field Lesions

Lesions at different points in the visual pathway produce characteristic field defects. A lesion of the optic nerve causes total blindness in the ipsilateral eye. An optic chiasm lesion results in bitemporal heteronymous hemianopsia (loss of both temporal fields). Optic tract lesions cause contralateral homonymous hemianopsia. Within the optic radiations, a temporal lobe lesion affecting Meyer's Loop causes a superior contralateral quadrantanopsia, while a parietal lobe lesion causes an inferior contralateral quadrantanopsia. Primary visual cortex lesions typically cause contralateral homonymous hemianopsia, often with macular sparing if the blood supply from the middle cerebral artery is intact, though the posterior cerebral artery is the primary supplier of V1. Higher-order agnosias are specific to the processing streams; ventral lesions can lead to Agnosias (Aperceptive or Associative), where patients cannot recognize objects or faces (Prosopagnosia), or perceive color (Achromatopsia). Dorsal lesions can lead to Akinetopsia (motion blindness), Optic Ataxia (improverished reaching), or Hemispatial Neglect.

Summary of Higher Visual Syndromes

In the ventral stream, aperceptive agnosia involves a failure in basic form reconstruction, while associative agnosia involves a failure to link a perceived object to its meaning (sémantique). Prosopagnosia specifically targets face recognition despite preserved recognition of facial features. In the dorsal stream, damage to MT/V5 results in motion blindness. Severe parietal damage can lead to Balint Syndrome, which includes ocular apraxia (fixed gaze), optic ataxia (visuomotor deficit), and simultanagnosia (the inability to perceive more than one object at a time). These deficits highlight the functional independence of the systems for identifying "what" an object is versus "where" it is and how to interact with it.