@5 on Thursday, 11 May, 2000
PVC
posterior striate cortex of macaque monkey
lamination pattern based on density of nissel staining
relate input to PVC to the lamination pattern
LGN
radioactive tracer into layers of LGN
parvocellular = transported up to cortex along axons � 4Cbeta (radioactivity visible in dark band)
magno � 4alpha
koniocellular pathway
complicates interpretation of m and p pathways
may be implicated in colour vision
probably there, to an extent in the old world primates (more difficult to visualise)
because can get different antibodies to stain the LGN � identify this 3rd group of cells
interlaminar zones
effect on interpretation of lesions � must also be affecting the interlaminar cells
anatomy of resolution of m and p pathways
parvo
numerically signficant
injection of horseradish peroxidase into a parvo axons � receptive field like an LGN cell, some colour specificity � then slice up the cortex to establish where it terminates
= quite tightly-focused for an affluent axon into the cortex
magno
much more diffusely spread
cone mosaic
simple cell � capable of responding 1-2 cones wide (= quite a focused input, as you would expect from the anatomy)
why do the bars seem to get more spaced apart on narrowing spatial frequency grating?
aliasing:
spatial frequency = too high, too fine for the pixels on the retina � appearing as lumpy stuff
some binoc stuff in the LGN � all inhibitory, not very precisely focused
mysteriously, there are more synapses to the LGN from the cortex than from the retina
very significant area of cortex = devoted to vision (25-50%)
Hubel + Wiesel � very old physiological recordings from V1
does work with spots, not just lines
disc covering whole receptive field � no response
can map out discrete receptive fields of simple cells
orientation + direction-selective
copmlex cell � does not have discrete on/off zones � light + dark are both able to excite receptive field
many cells for identifying constrast sign
but cortex � many cells which don�t care about the sign of the contrast
usually the contrast sign = very important (e.g. difficulty of recognising negative photograph)
moved recording site at 20micron intervals (averaged a number of cells, rather than single units) along 2mm of the surface of the cortex to show smooth curve of preferred orientation
binocular cells actually = equally balanced input from the left + right eyes
is not actually co-ordinated binocular input
all these experiments in anaesthetised animals
eyes were paralysed with curare-like drug so that they don�t wobble, but unfortunately this causes the eyes wander off in different directions
inject tracer into one eye
transports up through the LGN and into the cortex
correlate the anatomical readings electrophysiologically by penetrating in an electrode too
very tightly focused (deep) vs very wide speading cells (50microns across most of an ocular dominance band)
but we don�t really understand how many types of cells there are in the cortex
there is a lot of mixing of signals at this stage
blobs
probably getting specific input from the interlaminar koniocellular cells
reverse correlation
the stimulus sequence is presented, e.g. bar flashing on coputer screen at random locations
then, each time an AP occurs, refer it back to the stimulus on screen which preceded it by a set time
build up a picture of the receptive field through time, dynamic map for single cell of cortex of on/off regions
basic temporal response = biphasic (like a pendulum)
you can have end-stopped complex + simple cells
predictability � defines simple cells as simple
structure in the cortex � hypercolumn:
= a set of orientation-selective detectors for that particular location in the field
also ones with different scales, spatial frequency selectivities
would obviously need many more small than large receptive fields
give a functional interpretation to cellular properties
pipette intracellularly into a cell in the PVC
spikes selective for particular orientation, leads to much stronger depolarisation, much stronger than the noise
much discussion about: whetherh the selectivity for orientation primarily excitatory or inhibitory
old H+W theory: a row of on centre LGN cells �/span> simple cells
initial input = excitatory, but not very focused
much of the orientation selectivity = inhibitory
both EPSPs and IPSPs respond to the same region
\ H+W's theory cannot be right, since there is both EPSPs and IPSPs on the same on/off region within a receptive field
as opposed to EPSPs from all regions (whether on or off), but of different flavours
�/span> amplified, push/pull, differentiated
can�t visualise these simultaneously at the same membrane potential
nissel(?) staining
doughnut shaped receptive fields