A reinforcement learning framework for goal-directed stimulation of engineered neuronal networks

Systematically perturbing neuronal circuits under controlled conditions offers a promising route to understanding how they encode and process information, yet remains methodologically demanding. In vitro neuronal networks can be cultured on microelectrode arrays (MEAs) that allow stimulation and recording, and microfluidic patterning can constrain network topology to yield stable stimulation-evoked responses. Yet, the space of possible spatiotemporal stimulation patterns remains too large for exhaustive exploration. Additionally, the evoked responses depend on prior stimulation history. Here, we present an open-source framework that sustains closed-loop reinforcement-learning (RL) control of engineered neuronal networks, sending electrical stimuli to the MEA and evaluating the evoked responses to identify stimulation patterns that drive specific target activity motifs. We extend inkube, a low-cost, open-source electrophysiology system from off-the-shelf components, with closed-loop electrophysiology. This enables reliable delivery of stimulation at single-sample precision with millisecond-range round-trip times. It also allows independent RL agents to control multiple networks simultaneously. As a proof-of-principle application, we benchmarked different agents on the task of identifying stimulation patterns that maximise the length of clockwise-circular firing sequences. We first demonstrated that stimulation-evoked responses in engineered recurrent networks were stable and separable across the action space over hours of continuous operation. We characterised the dependency of responses on prior stimulation history, finding state dependence in a subset of stimulus pairs. The agents learned to improve reward during training with respect to random stimulation. Agents converged on non-trivial stimulation patterns that span the full action space rather than mirroring the target motif. All hardware designs and software are publicly available, providing an accessible platform for goal-directed functional characterisation of engineered neuronal networks at single-spike resolution.