BMI cyberworkstation: Enabling dynamic data-driven brain-machine interface research through cyberinfrastructure

Abstract

Dynamic data-driven brain-machine interfaces (DDDBMI) have great potential to advance the understanding of neural systems and improve the design of brain-inspired rehabilitative systems. This paper presents a novel cyberinfrastructure that couples in vivo neurophysiology experimentation with massive computational resources to provide seamless and efficient support of DDDBMI research. Closed-loop experiments can be conducted with in vivo data acquisition, reliable network transfer, parallel model computation, and real-time robot control. Behavioral experiments with live animals are supported with real-time guarantees. Offline studies can be performed with various configurations for extensive analysis and training. A Web-based portal is also provided to allow users to conveniently interact with the cyberinfrastructure, conducting both experimentation and analysis. New motor control models are developed based on this approach, which include recursive least square based (RLS) and reinforcement learning based (RLBMI) algorithms. The results from an online RLBMI experiment shows that the cyberinfrastructure can successfully support DDDBMI experiments and meet the desired real-time requirements.

Prerequesites for symbiotic brain-machine interfaces

“…In preliminary studies, we have quantified in brain control that animals can used their neural modulation and rewards that are projected on the external world to achieve control of a robotic arm [55]. The system involved the integration of improved real-time signal processing methods that capture global computation on multiple spatial, temporal, and behavioral scales [56,57]. …”

Model development, testing and experimentation in a CyberWorkstation for Brain-Machine Interface research

“…As shown in Fig. 3, multiple BMI models, such as RLS and RLBMI [1,2], and other service modules can be simultaneously connected to the experiment engine. This approach decouples all functionalities of the CW from each other and from the experiment engine, and allows these modules to be added or removed from the system with minimal changes to existing code of the experiment engine.…”

“…In our previous work [1,2], we have demonstrated how BMI control schemes (Recursive Least Square and Reinforcement Learning based BMI) can be implemented and tested in online and offline closed-loop experiments on the CW. Significant speed improvement in experiment execution has been observed when compared to the performance that can be achieved in a typical neurophysiology lab setting.…”

Cyber-Workstation for Computational Neuroscience

“…In an online RLBMI experiment with 8151 time steps (13 m 22 s), 99% of closed-loop control cycles were completed in less than 10 ms; 100% completed within 100 ms (Zhao et al, 2008 ). This demonstrates the CW provides a high-performance computing environment capable of real-time experiments including BMI adaptation.…”