What question did this study set out to answer?

To propose a new model of visual processing that emphasizes the retina's active role in information handling.

February 2, 2026Open Access

Retina-Centric Model of Visual Processing: Cascading Photon Transduction Without Cortical Image Reconstruction

Key Points

To propose a new model of visual processing that emphasizes the retina's active role in information handling.
Introduced the Retina-Centric Model (RCM) as an alternative to traditional cortical image reconstruction theories.
Analyzed metabolic constants and temporal dynamics in retinal and cortical processing.
Identified clinical observations supporting the RCM without relying on cortical compensation.
Demonstrated that the retina performs significant compression of visual information.
Found that the cortex primarily integrates pre-processed information rather than reconstructing images.
Predicted that visual processing aligns with metabolic efficiency principles.

Abstract

Current visual neuroscience proposes that the cortex reconstructs visual images from a compressed retinal input of 1.2 million nerve fibers transmitting ~10 Megabits/second. This framework creates a thermodynamic puzzle: cortical visual areas consume 4 Watts while retina consumes only 0.5 Watts, yet cortex is theorized to perform the computationally intensive task of image reconstruction. We propose the Retina-Centric Model (RCM): the retina is not a passive photoreceptor array but an active neural preprocessor that performs sophisticated 130-million-to-1.2-million compression via variable-speed photochemical cascades. The cortex does not reconstruct images; it integrates pre-processed information for binocular fusion, temporal coherence, and conscious access. RCM is grounded in:- Documented metabolic constants: ATP retinal = constant (~10⁸ ATP/s, Okawa et al. 2008); ATP cortical = constant (4W, Wong-Riley 2012)- Quantified temporal dynamics: Photochemical cascades accelerate 4-10× in milliseconds with light (Okawa et al. 2008)- Verified damage mechanisms: Retinal phototoxicity is photochemical (ROS-mediated, not thermal; literature consensus)- Measurable adaptation timescales: Gradual light changes (atardecer/amanecer) imperceptible across 20-60 minutes; abrupt changes (cinema/cave) perceptible in seconds- Falsifiable predictions: ATP constancy, ROS correlation with damage, cascade acceleration, cortical ATP independence from light The model explains clinical observations (rapid post-LASIK recovery, optogenetic acuity ceiling, myelination-constrained plasticity, amblyopia irreversibility) without invoking cortical compensation or image reconstruction. RCM suggests that visual system organization follows the principle of metabolic parsimony: delegate processing to the stage with lowest energetic cost per unit computation. Keywords: Retinal preprocessing, Photochemical cascades, Neural compression, Metabolic efficiency, Temporal adaptation, Photon transduction, Myelination constraints, Neuromorphic vision.

Read Full Paperexternally

Demander à l'IA

Bookmark

View Full Paper