Neurocomputing with brain organoids: closed-loop learning and network dynamics
Tuesday, April 28, 2026 | 17:00 CEST
08:00 PDT | 11:00 EDT | 23:00 CST | 00:00 JST
Webinar Hightlights
- How brain organoids can be used to study network dynamics and functional organization in vitro.
- What it takes to run closed-loop electrophysiology: reading, writing, and feedback-driven training of living networks.
- How HD-MEA functional phenotyping can quantify changes in activity patterns and connectivity over time.
- How organoid bio-model advances translate into biocomputing-style benchmarks and experimental pipelines.
- Practical considerations for implementing these workflows on MaxOne (organoid culture, longitudinal recording, stimulation, and closed-loop interfacing).
The webinar covered
- How brain organoids can be used to study network dynamics and functional organization in vitro.
- What it takes to run closed-loop electrophysiology: reading, writing, and feedback-driven training of living networks.
- How HD-MEA functional phenotyping can quantify changes in activity patterns and connectivity over time.
- How organoid bio-model advances translate into biocomputing-style benchmarks and experimental pipelines.
- Practical considerations for implementing these workflows on MaxOne (organoid culture, longitudinal recording, stimulation, and closed-loop interfacing).
Agenda
From Self-Organization to Learning: Mouse Forebrain Organoids as Models of Network Assembly and Goal-Directed Adaptation
Abstract
Brain organoids derived from pluripotent stem cells offer a powerful in vitro platform to study how neural networks self-organize and adapt. In this webinar, we present two complementary studies from our group at UC Santa Cruz. First, we describe the generation of dorsal and ventral mouse forebrain organoids and show how their distinct cellular compositions, particularly the enrichment of parvalbumin-positive interneurons in ventral organoids, give rise to divergent network topologies. Both organoid types develop small-world architecture without sensory input, but differ in modularity, hub organization, and synchronization dynamics, revealing how excitatory-inhibitory balance intrinsically shapes circuit formation (Hernandez, Schweiger et al., Stem Cell Reports 2026). Second, we demonstrate that cortical organoids can perform goal-directed learning when embodied in a pole-balancing task through closed-loop electrophysiology. Using adaptive training signals selected by reinforcement learning, organoids significantly improve their control performance in a manner which is shown to require intact glutamatergic transmission (Robbins et al., Cell Reports 2026). Together, these studies establish mouse forebrain organoids as a tractable system for investigating both intrinsic network assembly and stimulus-driven neural plasticity.
%2520Li.avif)