Vision III: Cortical mechanisms of First you tell them what youre - PDF document

Vision III: Cortical mechanisms of First you tell them what you’re vision gonna tell them Please sit where you can examine a • The cortical visual system is composed of multiple visual areas with different functions. partner. • V1 neurons describe object features. • The principle of columnar organization. • Two visual streams – ‘what’ and ‘how’ (or ‘where’). Michael E. Goldberg, M.D. • MT neurons describe motion and depth (dorsal stream). • IT neurons describe objects (ventral stream). See the triangle? See the white bar? See the wavy line? Which small square is darker? 1

Roughly 40% of cerebral cortex is So involved in vision • Your visual system does not measure and report the exact physical nature of the visual world. • It collects some data, and makes guesses. • Optical illusions take advantage of the guessing strategies. The striate cortex – V1 – builds more Remember sophisticated receptive fields from these basic building blocks. Cells describe specific • Receptive fields in the retina and the • Contour orientations. lateral geniculate are circular, with • Binocular interaction. center – surround organization. • Speed and direction of motion. Off surround - inhibits • Color. On center - excites V1 simple cell is most responsive to Orientation tuning in a V1 simple an oriented line cell Off-response Spikes/second On-response Stimulus Angle (from max) 2

V1 complex cells are sensitive to But not particularly to stimulus orientation of stimuli position within the receptive field Complex cells can be constructed The cerebral cortex is organized in from an array of similarly oriented a columnar manner simple cells geniculate nucleus 1 2 Blob Pyramidal Pyramidal 3 4A 4B 4C α 4C β Spiny stellate Smooth 5 (local) stellate (local) 6 To extrastriate To SC,pulvinar To LGN, I P P P M M Cortex – V2,V3 pons claustrum V4, MT Cells with similar orientation Within a column preferences lie in the same column • Information is processed and transformed from monocular, center-surround,non-directionally selective input to • Orientation- • Binocular disparity- • Direction-selective output • Processed information is distributed • Layers 2-3 to other cortical areas • Layer 5 to the superior colliculus • Layer 6 to the lateral geniculate nucleus • This general arrangement of columnar processing is maintained throughout the cortex, not just visual cortex. 3

Geniculate cells representing the same Orientation columns with the same area of the visual field but arising from monocular lateral geniculate input lie in different eyes project to adjacent areas the same ocular dominance column. of V1 Color sensitive cells lie at the center of the The actual topology of orientation pinwheels, in cytochrome oxidase containing and ocular dominance columns ‘blobs.’ Color sensitive cells are mostly unoriented Depth perception starts with the Random dot stereograms generate detection of binocular disparity structure from disparity B C A A L A R C B L B R C 4

Disparity selectivity in a V1 neuron Motion selectivity in a V1 neuron Two cortical visual streams Patients demonstrate this functional subserve two different visual segregation functions. • Patients with V1 lesions generally have total Where/how? visual field deficits in the affected field. • Patients with dorsal stream lesions have deficits in sensory location (and attention), motion perception, color perception, and the performance of visually-guided movements. • Patients with ventral stream lesions have visual agnosia, the inability to associate a What visual stimulus with a name or function. Functional separation begins in the retina and continues through the And continues in V1 LGN LGN Parvocellular cells Interblob Blob Retinal P cells: color, longer latency, fine detail Retinal M cells: LGN Magnocellular cells broadband, shorter latency courser detail 5

V2 (Area 18) also is divisible by Functional separation continues in cytochrome oxidase staining V2 Stripes in Area 18 Blobs in Area 17 After V2, different functions are performed by After V2, different functions are performed by anatomically different areas: anatomically different areas: The dorsal stream provides vision for action –”where The ventral stream provides vision for object and how” identification After V2, different functions are performed by anatomically different MT – the analysis of motion areas: But the areas are interconnected • Neurons in MT are selective for speed and direction of motion, and retinal disparity. • Neurons in MT report the perceptual aspects of motion. • Electrical stimulation of MT affects the perception of motion. 6

Human MT Structure from motion MT Cells are tuned for direction Perceived motion in a plaid Striate neurons respond to the MT responds to the direction of the components of the plaid plaid, and not the components Single component Plaid (2 components) Single component Plaid (2 components) 7

MT has columns for direction of MT has disparity columns motion Electrically stimulating an orientation Electrically stimulating an orientation column in MT induces the perception of column in MT induces the perception of motion described by that column motion described by that column 100% coherence 50% coherence Electrically stimulating an orientation Electrically stimulating an orientation column in MT induces the perception of column in MT induces the perception of motion described by that column motion described by that column No coherence 8

The parietal lobe describes the There are multiple representations world for action, location, and of the visual field in the intraparietal attention. sulcus Where/how? What Within the dorsal stream there is An example of a dorsal stream further functional segregation – function • MT is specialized for depth and motion. • When you reach for something, your grip opens to accommodate the size of • LIP is specialized for attention in far space. your target. • MIP is specialized for providing visual. • Patients with dorsal stream lesions can’t information for reaching. do this. • AIP is specialized for providing visual. • They can, however, describe the size of information for grasping. the object. • VIP is specialized for providing visual. information for mouth and head movements A patient with a dorsal stream lesion cannot orient her hand with respect to a Neurons in AIP specialized for grip slot Look at the Reach for object the object 9

The inferior temporal lobe describes A patient with a ventral stream lesion the visual world for object can move her hand to a slot, but can’t mimic the position. recognition Matching Posting Where/how? DF What? Control Neurons in inferior temporal cortex are selective for complex patterns like faces Patients with inferior temporal Ventral stream patients lesions have visual agnosia • Cannot identify objects Copy the drawing • But they can make appropriate visually- Visuomotor function Intact – but patient guided movements. can’t name the object • The patient who could not set her grip can still tell you which cylinder is thicker. Draw an anchor. Patient cannot • The patient who cannot tell you which conceptualize the cylinder is thicker can set still her grip. anchor 10

Prosopagnosia “face blindness” is Finally, you tell them what you told the most dramatic ventral stream them deficit • The striate cortex (V1) uses unoriented, monocular • Term first used by Bodamer, 1947 input from the lateral geniculate to assemble cells • Inability to recognize familiar faces selective for orientation,motion, and retinal disparity. Complex cells generalize the orientation information • Visual acuity is normal found in simple cells. • Striate cortex is organized in columns with similar • Caused by lesion to right inferior temporal orientation and ocular dominance. lobe • Two visual streams emanate from V1: a dorsal stream concerned with analyzing the visual world for • May be congenital (“developmental location and action, and a ventral stream concerned prosopagnosia”) with analyzing the nature of objects in the visual world. Different areas subsume different spatial and • Patients compensate by using other object attribute functions. recognition cues: clothing, gait, voice, etc. • Clinical deficits include specific deficits for color, faces, motion, visual targeting of motion, and spatial 11