BRAID: Input-Driven Nonlinear Dynamical Modeling of Neural-Behavioral Data
This work addresses the challenge of accurately modeling neural-behavioral relationships in neuroscience, particularly for motor tasks, but it is incremental as it builds on existing input-driven recurrent neural network methods.
The authors tackled the problem of modeling neural dynamics without accounting for external inputs by introducing BRAID, a deep learning framework that incorporates measured inputs to disentangle intrinsic dynamics from input effects, resulting in more accurate fitting and improved forecasting of neural-behavioral data compared to baselines.
Neural populations exhibit complex recurrent structures that drive behavior, while continuously receiving and integrating external inputs from sensory stimuli, upstream regions, and neurostimulation. However, neural populations are often modeled as autonomous dynamical systems, with little consideration given to the influence of external inputs that shape the population activity and behavioral outcomes. Here, we introduce BRAID, a deep learning framework that models nonlinear neural dynamics underlying behavior while explicitly incorporating any measured external inputs. Our method disentangles intrinsic recurrent neural population dynamics from the effects of inputs by including a forecasting objective within input-driven recurrent neural networks. BRAID further prioritizes the learning of intrinsic dynamics that are related to a behavior of interest by using a multi-stage optimization scheme. We validate BRAID with nonlinear simulations, showing that it can accurately learn the intrinsic dynamics shared between neural and behavioral modalities. We then apply BRAID to motor cortical activity recorded during a motor task and demonstrate that our method more accurately fits the neural-behavioral data by incorporating measured sensory stimuli into the model and improves the forecasting of neural-behavioral data compared with various baseline methods, whether input-driven or not.