The Vision System and the Brain I
Why is vision special?
• For humans, who are diurnal, vision is vital
sense
• Can perceive information at a distance
• Localise objects accurately
• Vision tends to dominate over other senses
• Approx 1/3 of cerebral cortex dedicated to
visual analysis and perception
• Most widely researched sensory system
Visual system: complex processor
• The visual system has to:
– translate discrete points of light falling onto our
photoreceptors in the retina into meaningful
objects that we recognise
– discriminate objects from other aspects of the
visual scene i.e. background environment
– recognise these objects in different orientations,
even if it has only ever seen a lion from the front
• Today: how the brain begins to represent
sensory information & processes it to form
integrated percepts
Parallel processing in the visual system
• What do neural signals in the visual
pathway represent?
• The central hypothesis is that visual
information is distributed across
subsystems
• Consequently, perception is analytic –
some processes represent shape,
others colour, movement etc.
We perceive objects as unified wholes – but how?
We do not have the impression that our perception of the red car driving past was created
in a piecemeal fashion (I.e. the colour, motion and shape) are processed separately.
However, evidence from cognitive neuroscience provides compelling support for the idea
that perception DOES operate in an analytic manner – this is called the feature-extraction
hypothesis.
Visual information is contained in the
light reflected from objects
As light passes through the lens, the
image is inverted and an optical image is
focused onto the back surface of the eye,
the “retina”.
The eye as a camera
• To produce optimal images, eye has many features
of a camera – e.g. cornea and lens to focus light
• Light travels in a straight line until it hits a change in
the density of the medium which causes light rays to
bend (refraction)
• The cornea has a fixed curvature and bends light
rays – responsible for forming the image on the retina
• Focus is adjusted by changes in the shape of the lens
controlled by ciliary muscles inside the eye
• Myopia (short-sightedness) results from steeper
cornea/longer than normal eyeball, so image is
focused in front of the retina
Visual processing begins in the retina
Receptor cells are highly specialised, and sense selective changes in the
environment, e.g., light (photons) and sound waves. Stimulation leads to
changes in receptor molecules, which open/close ion channels.
Vision: photoreceptors on the
retinal surface contain pigments
that are sensitive to light
Photoreceptors: rods and cones
Photoreceptors contain specialised membranes with pigment molecules for
phototransduction. There are two types of photoreceptors:
(i) Cones (100M) contain 3 pigments and contribute to high acuity colour and daytime vision. Photopic system (Gk photos light) Fovea 1
(ii) Rods (10M) contain rhodopsin, which is very sensitive, and contribute to night
vision. Scotopic system (Gk skotos darkness)
Spectral sensitivity functions for rods + cones
Retinal cells sense
light information from
a specific point of the
world (receptive
field). Rods and
cones respond to
point of light in a
particular range of
wavelengths.
Anatomy of the retina
Changes in Glutamate release in photoreceptors cause graded currents in
bipolar cells. Stimulation of ganglion cells by bipolar cells results in action
potentials, which travel through the optic nerve into the brain.
Photoreceptors: pigments and receptive fields
rods and cones
ganglion cells
Retinal cells sense light information from a specific
point of the world (receptive field). Rods and cones
respond to points of light in a particular range of
wavelengths.
Inputs from photoreceptors are integrated into
opponent center-surround RFs in ganglion cells.
Optic nerve
The axons of the ganglion cells form the optic nerve which carried visual signals
from the retina into the brain. Some axons cross at the optic chiasm. The result is
that each visual field is represented contralaterally (opposite side) in the brain.
Visual receptive fields (RF’s)
• The whole area of the world that you can see
at any one time is called your visual field
• The part that you see to your left is your left
visual field, and the space on the right is the
right visual field
• The part of the visual field to which any one
neuron responds is that neuron’s receptive
field.
Concentric
receptive fields
Ganglion cells (like those in
the LGN further along the
visual pathway) have
concentric receptive fields with
either a centre-off or centre-on
surround. Cell fires rapidly
when light hits centre, is
inhibited when light is over the
surround, and halfway across
–no activity.
As such, these cells are ideal
for signalling changes in
illumination (contrast) such as
those that arise from stimulus
edges.
Overview of visual transduction
• Output from photoreceptors is first
processed in bipolar cells and from
there to the ganglion cells.
• 260 million photoreceptors and only 1
million ganglion cells – so already
substantial compression of information
• Axons of the ganglion cells form a
bundle call the optic nerve
Lateral geniculate nucleus
Visual signals from the retina into the LGN
are segregated according to (1) the type of
ganglion cell and (2) the eye of origin.
The LGN contains 6 layers of cells. The
inner 2 contain large cells (magnocellular)
and the outer 4 contain small cells
(parvocellular). In between layers
(interlaminar) are other small cells
(koniocellular). These layers receive input
from different types of ganglion cells.
Input from each eye alternates within these
types of layers.
Receptive fields in the LGN are similar to
those in the retina.
Lateral
geniculate
nucleus
(LGN)
90% fibres from optic
nerve terminate in
LGN. Projections are
not random but rather
terminate selectively
in one of the six layers
that comprise LGN.
Ipsilateral eye projects
to layers 2, 3, 5.
Contralateral eye
projects to 1, 4, 6.
Layers 1&2 :
magnocellular system.
Anatomy of the visual system
In the main
pathway, retinal
signals travel to the
lateral geniculate
nucleus (LGN) of
the thalamus and
then to primary
visual cortex (V1).
There are also
alternative
pathways, e.g.,
through the
superior colliculus
(SC).
Before entering the
brain, each optic
nerve splits into 2
parts
Three kinds of cells in LGN
Parvocellular neurons
Form “P” Pathway
Magnocellular
neurons
Form “M” pathway
Koniocellular
neurons
Cell bodies
Smaller
Larger
Small
Receptive fields
Smaller
Larger
Mostly small,
variable
Retinal location
In and near fovea
Throughout the retina
Throughout the
retina
Colour sensitive
Yes
No
Some are
Respond to
Detailed analysis of
stationary objects
Movement and broad
outlines of shape
Varied and not yet
fully described
Primary visual cortex (V1)
The first cortical synapses for neurons
carrying visual information are in the
medial portion of the occipital lobe –
Area 17 in Brodmann’s map.
This area is located medially (midline)
and buried below the superficial layer of
the cortex along the calcarine sulcus.
The cytoarchitecture is quite regular and stippled – thus giving rise to the name
“striate” cortex. Also referred to as primary visual cortex. This name reflects the
hypothesis that this is the first visual processing area in the cortex.
Anatomy of V1
The segregation of the M and P pathways is maintained in the cortex. Axons from both regions terminate in layer 4
of the striate cortex but (a) the terminal zones of axons in this layer are offset from one another and (b) p pathway
involves a 2nd synapse carrying information into the more superficial layers 2 and 3.
V1: blobs and interblobs
Different types of inputs to V1 form a pattern of
“blobs” revealed by staining for cytochrome oxidase
(mitochondrial enzyme). Blobs extend through
layers 2, 3 and (less so) 5, 6.
Parvocellular and koniocellular inputs linked to
colour processing form blobs.
Parvocellular inputs concerned with shape analysis
and magnocellular inputs form interblob regions.
Anatomy of the visual system
In the main
pathway, retinal
signals travel to the
lateral geniculate
nucleus (LGN) of
the thalamus and
then to primary
visual cortex (V1).
There are also
alternative
pathways, e.g.,
through the
superior colliculus
(SC).
Before entering the
brain, each optic
nerve splits into 2
parts
V1 - Columnar organisation
Neurons with similar selectivities cluster in columns, according to dominant eye and
preferred orientation. Organisation revealed by 2DG and optical imaging. Orientation
columns regularly interrupted by blobs.
Hypercolumns (cortical modules) contain full set of orientation, for both eyes.
Occular Dominance Columns
Orientation Selectivity
V1 - Blindsight
DB—patient at UL—had
migranes for years.. and was
found to have a tumour on left
side of V1 which was removed
which reduced migranes but
also left him apparently
BLIND in right visual
pathway.
Wieskrantz—tested DB..and
he could see on his right VF!
Lesions to V1 result in Blindsight. The person is phenomenologically blind. However,
many residual unconscious visual functions remain. The individual can point to stimuli
s/he cannot see, and guess about many aspects of the stimulus with performance
that is much better than chance.
Blindsight illustrates the existence of parallel circuitry for the processing of visual
information that can guide action. It also raises interesting questions regarding the
nature of visual awareness.
Extrastriate visual cortex
Visual information in extrastriate cortex forms 2 largely independent parallel streams.
Ventral stream (to inferior temporal) computes stimulus appearance (“what”) and
contains input from parvocellular, koniocellular and magnocellular cells.
Dorsal stream (to posterior parietal) computes stimulus location and movement (“where”)
from magnocellular inputs mainly.
Ventral visual areas
IT face-selective cells
achromatopsia
V4 is selective for colour and form. RFs large and more complex than blobs (V1) and
thin stripes (V2). Neurons respond to many wavelengths and compensate for illumination
(colour constancy). Lesions can contribute to achromatopsia.
IT is downstream from V4. RFs are huge, selective for objects or complex shapes
(*faces), and invariant to location or backgrounds. Lesions ~ agnosia or prosopagnosia
(object- or face-recognition deficits).
Neuronal activity modulated by attention and memory.
What does an agnosic experience?
Dorsal visual areas
akinetopsia
hemispatial neglect
MT is selective for motion. RFs large and more complex than interblobs (V1). Lesions
contribute to akinetopsia (motion blindness) .
PPC is downstream from MT. RFs are large and map spatial location along different
reference frames for different types of movements. Lesions result in Balint’s syndrome
(bilateral occipital-parietal lesions) or hemispatial neglect (inferior parietal lobe).
Neuronal activity modulated by attention and memory.
W. W. Norton
In summary …
Vision is one example of transducing energy into perception
Vision exemplifies general scheme for perception
periphery>thalamus>primary sensory area
>associative unimodal areas… (feedback)
Visual system exemplifies functional specialisation
starts from sensory receptor cells
occurs within brain regions (retina, LGN, V1)
occurs between brain regions (V4, IT, MT, PPC)
occurs between brain circuits (ventral, dorsal)
Visual system also exemplifies concurrent processing
massive anatomical interconnectivity
residual abilities survive perceptual deficits