Connectome Scanning: Digitizing a Brain

Scan Parameters

Voxel resolution 4 nm

Neuron density 1200

Scan speed 1.0×

Scan Progress

0% scanned0 / 0 neurons

Data (this volume)

0 GB

Reconstruction fidelity

—

Extrapolated to a full human brain (~1200 cm³)

—

How it works

"Brain upload" starts as a physical scanning problem: to reconstruct a connectome, tissue must be imaged at a resolution fine enough to resolve individual synapses (~20 nm clefts, thinner dendritic spine necks). This simulator models that scan as a plane sweeping through a synthetic neuron/synapse graph, and computes the real data-volume trade-off:

data(R) = D₄ · (4 nm / R)³ × V

D₄ ≈ 1.4 PB / mm³ at 4 nm resolution
(the published scale for electron-microscopy
connectomics, e.g. the 2021 Google/Harvard
"cubic millimeter of brain" reconstruction)

Because voxel volume scales as R³, halving the resolution multiplies the data by 8×. At 4 nm — enough to trace individual axons and see synaptic vesicles — a full human brain (~1,200 cm³) would need roughly a zettabyte of storage; coarsen the resolution and the volume shrinks fast, but so does the ability to reconstruct real synaptic connections.

Resolution slider — smaller nm/voxel = slower scan, more data, higher fidelity.
Density slider — how many neurons populate the modeled sub-volume.
Fidelity — an illustrative curve: near 100% once resolution reaches synapse scale (≤4–10 nm), falling off sharply above ~20 nm where synaptic clefts become unresolvable and connectivity is guessed rather than measured.

This is the scanning half of the "mind upload" idea taken literally: even before asking whether a copied connectome preserves identity, physically acquiring it at sufficient fidelity is already a multi-exabyte engineering problem with today's electron-microscopy throughput.

Unscanned tissue Digitized neuron Scan plane

⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.