A living organ in a Petri dish reveals loops express

Original: "Emergent topological structure in spontaneous brain-organoid activity"
arXiv:2607.16517 · 2026-07-17 · CC BY 4.0 · 1 min
Stable closed chains of neurons found in the spontaneous activity of brain organoids.

📄 Showing the "Simple" version — "Popular" is not ready yet. Add it to favorites to help prioritize it.

A city at night: lights on in some windows, dark in others. If you connect with lines the windows that light up simultaneously, you get a web of connections. Researchers studied such a web in tiny brain organoids — living clusters of nerve cells grown in the lab.

They recorded electrical signals from dozens of neurons and built a network where the strength of the connection between cells depended on how synchronously they fired. Then, using a mathematical method, topological analysis, they searched this network for loops: closed chains of three or more neurons. A loop is not just a group of friends but a structure capable of transmitting a signal in a circle.

Loops were found in 14 out of 18 experiments. They were too frequent to be a coincidence.

These loops turned out to be surprisingly resilient. Random removal of neurons barely destroyed them. But when they specifically removed the cells that were part of the loops, the structure fell apart. As the organoid grew, more complex forms also appeared — voids in the network, resembling cavities in a sponge.

Thus, in the electrical noise of living neurons, a clear geometric framework emerged. The same loops that flickered in the network are now visible as a basis for future rhythms and waves, like city lights forming a stable pattern.

Brain organoids are grown from stem cells, and they can spontaneously generate electrical impulses reminiscent of the activity of a real brain.

Scientists
Andrey Markov
Concepts
neural network memory architecture topology hippocampus random network information storage microelectrode array brain connectivity stochastic self organization working memory neural recording technology entropy driven self assembly computation topological pattern formation electroencephalography Hurst exponent band structure learning brain neuron Recurrent Neural Networks