Inferring brain plasticity rule under long-term stimulation with structured recurrent dynamics

A research team led by Zhichao Liang has introduced the Stimulus-Evoked Evolution Recurrent dynamics (STEER) framework, a breakthrough approach for understanding how long-term brain stimulation reshapes neural circuits. While short-term synaptic changes are well-studied, the principles governing circuit-level reorganization across weeks remained unknown until now. STEER formalizes these principles as a latent dynamical law that governs how recurrent connectivity evolves under repeated interventions, elevating long-term plasticity from a hidden confound to an identifiable dynamical object that can be optimized.

The STEER framework operates on dual timescales, separating fast neural activity from slow plastic changes and representing plasticity as low-dimensional latent coefficients evolving under learnable recurrence. The team validated STEER across four benchmarks including synthetic Lorenz systems, biologically grounded BCM-based networks, task learning with optimized stimulation, and longitudinal recordings from Parkinsonian rats receiving closed-loop deep brain stimulation (DBS). Results demonstrate that STEER recovers interpretable update equations, predicts network adaptation under unseen stimulation schedules, and supports designing improved intervention protocols. This provides a data-driven foundation for both mechanistic insight into brain function and principled optimization of therapeutic brain stimulation.