PSYC 210 — Exam 2 Guide (F26 UIUC)
Retina and Photoreceptors
Photoreceptor Distribution across the Retina:
The fovea is characterized by a high density of photoreceptors consisting almost exclusively of cones, optimizing visual acuity for high-detail, central vision.
The retinal periphery consists predominantly of rods, which provide high sensitivity for vision in low-light conditions.
Tristimulus Cone Architecture:
Human color vision relies on three distinct classes of cone photoreceptors categorized by their spectral sensitivities:
L-cones: Long-wavelength sensitive.
M-cones: Medium-wavelength sensitive.
S-cones: Short-wavelength sensitive.
Activation by white light stimulates all three cone classes in approximately equal proportions, represented as the vector .
The Principle of Univariance:
An individual cone cannot independently encode or differentiate color.
A single photoreceptor's hyperpolarization response depends jointly on the wavelength and the intensity of the incoming photons, making it impossible to deduce wavelength from the output of a single receptor alone.
Retinal Output and Ganglion Cells:
Retinal ganglion cells (RGCs) are the sole neurons in the retina that project axons out of the eye into the central nervous system (CNS).
ON-center retinal ganglion cells increase their action potential firing rate in response to light stimulation falling directly in the center of their receptive field.
Bipolar Cell Neuromodulation:
Bipolar cells do not share a uniform response to the neurotransmitter glutamate.
Glutamate produces differential signaling based on cell type:
ON-bipolar cells are hyperpolarized (inhibited) by glutamate via metabotropic receptors.
OFF-bipolar cells are depolarized (excited) by glutamate via ionotropic receptors.
Receptive Fields and Neural Coding Principles
Receptive Field Foundations:
A receptive field is defined as the specific sensory region (such as a patch of the visual field or sensory surface) where stimulation causes a change in a neuron's firing rate.
Receptive field sizes expand dramatically across the visual hierarchy, ranging from exceedingly small receptive fields in retinal ganglion cells to exceptionally large receptive fields in the inferotemporal (IT) cortex.
Tuning Curves:
A tuning curve provides a quantitative graphical representation of a neuron's average firing rate plotted as a function of continuous variation in a specific stimulus parameter (e.g., orientation, spatial frequency, wavelength, or motion direction).
Neural Coding Schemes:
Rate Coding: Information is represented by the average number of action potentials (spikes) discharged per unit time.
Temporal Coding: Information is carried within the millisecond-precise timing or temporal interval patterns of individual spikes.
Labeled Line Coding: The qualitative identity and perceptual meaning of a stimulus are defined explicitly by which specific neuronal line or population fires, rather than how it fires.
Sensory Adaptation:
Prolonged, sustained sensory stimulation results in sensory adaptation, leading to a measurable decline in neuronal firing rate and reduced sensitivity to that continuous stimulus.
Visual Cortex Hierarchy and Receptive Field Scaling
Hierarchical Progression of Receptive Field Size:
Receptive field sizes systematically expand at each successive stage of visual processing:
Primary Visual Cortex (V1):
V1 is the earliest visual area where pronounced orientation tuning emerges.
Complex cells in V1 exhibit selective firing to oriented edges, bars, or contours while remaining invariant to the exact spatial position (phase) of the stimulus within their receptive field.
Secondary Visual Cortex (V2):
V2 contains specialized neurons dedicated to resolving border ownership (determining which side of an edge belongs to a foreground figure versus background ground) and selectivity for visual textures.
Visual Area V4:
Neurons in V4 process intermediate visual representations, exhibiting tuning for contour curvature and mid-level geometric 2D/3D shapes.
Inferotemporal (IT) Cortex:
IT cortex sits at the apex of the ventral stream object-recognition pathway, exhibiting high-level category selectivity (e.g., selective activation to complex objects, animals, and faces).
Mathematical Modeling in Vision
Mathematical Model of Cone Responses:
The response of a cone photoreceptor is mathematically formalized as the inner product (dot product) of the incident illumination's spectral power distribution and the cone's spectral sensitivity function.
Linear Receptive Field Modeling:
A linear receptive field is modeled mathematically as the dot product between an input sensory stimulus vector and the spatial weighting matrix (receptive field weights) of the neuron.
Bayesian Inference and Normative Models
Neural Stochasticity:
Biological neural responses are inherently noisy, exhibiting trial-to-trial variability even under identical sensory stimulation.
Bayesian Theoretical Framework:
Prior (): The baseline probability distribution representing the observer's expectations, prior knowledge, or evolutionary biases before any sensory evidence is gathered.
Likelihood (): The probability of observing the specific incoming sensory data given a particular state of the world or physical parameter .
Posterior (): The updated belief or perceptual inference, which is proportional to the product of the prior and the likelihood:
Influence of Sensory Noise on Perception:
When measurement noise increases, sensory certainty declines, flattening the likelihood function.
Higher measurement noise causes the posterior distribution to shift toward the prior, forcing perception to rely more heavily on prior expectations.
Normative Models of Neural Function:
Normative models explain biological neural architectures and computational processes by demonstrating that they represent optimal, mathematically formalized solutions designed to maximize behavioral performance or efficiency given environmental constraints and noise.
Motor System, Somatosensory Cortex, and Neurological Disorders
Primary Motor Cortex (M1):
Lesions, trauma, or physical damage to the primary motor cortex result in motor paralysis corresponding to the somatotopically mapped region of injury.
Somatosensory Cortical Topography:
The somatosensory cortex exhibits cortical magnification, allocating disproportionately large cortical territory to the hands relative to their physical surface area, reflecting fine tactile sensitivity and manual dexterity.
Parkinson’s Disease:
Parkinson’s disease is a neurodegenerative motor disorder caused by the progressive loss and degeneration of dopaminergic neurons in the substantia nigra of the basal ganglia.
Mirror Neuron Networks:
Mirror neurons are sensorimotor cells that fire action potentials both when an individual performs a goal-directed motor action and when the individual passively observes another agent executing that same action.