ISETBIO

ISETBIO simulates the early stages of human visual processing by modeling photon emission from displays, passage through ocular optics including fixational eye movements, cone photoreceptor isomerizations across wavelengths, and classification of stimulus orientation using machine learning.

Key Features:

• Matlab-based implementation: The toolbox is implemented in Matlab for computational modeling of visual optics and photoreceptor responses.
• Photonic interaction simulation: Simulates photons emitted from a display source and their passage through human ocular optics, including fixational eye movements.
• Cone isomerization modeling: Models wavelength-dependent absorption by retinal cone photoreceptors and the resulting isomerizations that underlie phototransduction.
• Psychophysical task simulation: Simulates orientation discrimination tasks and allows analysis of performance across different polar angles of the visual field.
• Machine learning integration: Employs a support vector machine (SVM) to classify stimulus orientations based on photon absorption patterns.

Scientific Applications:

• Optical quality and cone density analysis: Investigates how optical quality and cone density contribute to variations in visual performance across polar angles.
• Asymmetry investigation: Studies asymmetries in visual performance that are not fully explained by optical properties alone.
• Quantification of early-stage contributions: Quantifies the extent to which early visual processing stages account for differences in contrast sensitivity and orientation discrimination across meridians.

Methodology:

The computational workflow explicitly models photons emitted from a display, their propagation through simulated human optics including eye movements, wavelength-dependent cone photoreceptor isomerizations, and uses a support vector machine to classify orientation from the resulting absorption patterns.