Efficient coding matters in the organization of the early visual system.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Evolutionary Biology

Background:

• Brain areas form hierarchical networks shaped by evolution.
• Previous research suggests receptive fields (RFs) evolved for efficient neural coding.
• The organization of neurons and network hierarchy may also support efficient coding.

Purpose of the Study:

• To investigate if neural organization and hierarchical structure evolved for efficient coding.
• To simulate the early human visual system to test this hypothesis.

Main Methods:

• A three-layer feed-forward network with intra-layer connections was developed.
• The network simulated early stages of the human visual system.
• Neurophysiological data was used for comparison.

Main Results:

• Simulated receptive fields (RFs) for retinal ganglion cells (RGCs), lateral geniculate nucleus (LGN), and V1 simple neurons matched neurophysiological data.
• Closer RGCs showed correlated responses, and V1 simple neurons with similar orientations clustered.
• Hierarchical organization of the early visual system demonstrated energy savings.

Conclusions:

• The study supports the hypothesis that neural organization and hierarchical structure evolved for efficient coding.
• The findings align with the principles of metabolically efficient neural coding in human brain evolution.
• The model highlights the importance of intra-layer connections in visual system simulations.