The Thalamus and Primary Visual Cortex

Core Concepts

From retina to LGN: keeping the map intact

Retinal ganglion cell axons travel in the optic nerve to the optic chiasm, where fibers from the nasal (inner) half of each retina cross to the opposite side while temporal (outer) fibers stay ipsilateral. Each optic tract therefore carries the contralateral visual field — the left tract carries the right visual field, and vice versa. This partial decussation is why a chiasm lesion produces bitemporal hemianopia (loss of the outer halves of both fields), while a tract or cortex lesion loses only one half-field.

The tract terminates in the LGN, a layered structure. In primates the LGN has six principal layers: the inner two are magnocellular (M pathway: motion, low spatial frequency, fast conduction), the outer four are parvocellular (P pathway: fine detail, color), with thin koniocellular layers between them. Each layer receives input from one eye only, so the two eyes' signals are kept in separate processing channels. Retinotopy is preserved: neighboring retinal points activate neighboring LGN cells.

The LGN: relay and gate

The LGN has long been called a relay, but it is more than a passive switchboard: it receives far more input from the cortex (V1 feedback) than from the retina, plus modulatory inputs from the brainstem. Proposed roles include gating attention, coordinating signals with eye movements, and synchronizing activity. For behavioral neuroscience, the key point is that the M/P/K separation established in the retina is maintained through the LGN and delivered to distinct cortical targets — a processing architecture, not just a cable.

V1: the striate cortex and its layers

The primary visual cortex (Brodmann area 17) sits in and around the calcarine sulcus of the occipital lobe. It is called striate cortex because of a visible stripe (the line of Gennari) formed by dense myelinated axons in layer 4. Inputs arrive in layer 4:

• Magnocellular axons target layer 4Cα
• Parvocellular axons target layer 4Cβ
• Koniocellular input arrives in the superficial layers (and layer 1), feeding the blobs — cytochrome-oxidase-rich patches in layers 2/3 that process color.

From layer 4 the signal is recombined: neurons in other layers respond to oriented edges, motion, and color — features that did not exist in the retina's center–surround fields.

Simple and complex cells: the feature detector hierarchy

Hubel and Wiesel described two canonical V1 cell types:

• Simple cells: respond to an oriented bar or edge at a specific position within their receptive field. Their fields have distinct ON and OFF subregions arranged side by side, as if several LGN center–surround fields were lined up — a model that successfully predicts simple-cell responses.
• Complex cells: also orientation-selective but respond to the oriented stimulus anywhere within a larger field and often respond best to motion in a preferred direction. They appear to pool the outputs of simple cells.

V1 neurons are thus edge and contour detectors tuned for orientation, the raw materials from which object shape is assembled.

Columns, hypercolumns, and cortical magnification

V1 is organized as a series of repeating modules:

• Orientation columns: cells with the same preferred orientation sit in vertical slabs; neighboring columns shift preference systematically, covering all orientations in a regular progression.
• Ocular dominance columns: alternating bands of cells respond preferentially to one eye or the other.
• A hypercolumn is the full set of columns — all orientations, both eyes' dominance — serving one small point of visual space. It is the cortex's basic processing unit, repeated across the map.
• Cortical magnification: the fovea, which occupies a tiny fraction of the retina, commands a disproportionately large area of V1. This matches the fovea's high acuity and explains why central vision dominates perception.

Common Confusions