- Organisation
- Princeton University
- Year
- 2026
- Substrate
- 3D cultured neural network
- Interface
- 3D flexible electrode array
- Task
- Reservoir computing + chronic stimulation
- Evidence
- Peer reviewed
3D-MIND is a three-dimensional micro-instrumented neural network device reported by Princeton researchers in Nature Electronics in 2026. A flexible electronic array is folded through a three-dimensional culture of rat hippocampal neurons, allowing recording and stimulation from multiple planes through the network.
The authors report action-potential recording over six months, evolving connectivity maps, pharmacological response measurements and chronic electrical stimulation. They also use the network as a reservoir-computing substrate after tuning connectivity strengths.
Why the geometry matters
Planar electrode arrays provide excellent access to neurons that sit close to the surface but become less informative as biological networks acquire depth. Organoids solve part of the biological geometry problem while making electrical access harder. 3D-MIND attacks the interface problem directly by distributing electronics through the culture.
What it establishes
The strongest result is an interface result: a living three-dimensional neural network remained electrically accessible across depth for months while supporting repeated stimulation and computational experiments. The study advances dense bidirectional interfacing; general-purpose processor performance remains unestablished.