Brain-machine interfaces researches in rats

To study decoding information in the neural system, invasive brain-machine interfaces (BMIs) in rat were investigated. In this paper we developed a remote control training system for rat navigation. The system consists of an integrated PC control program with functions of signal communication, parameter setting and data file management, a transmitter and a receiver based on Bluetooth modules, and a stimulator controlled by a microprocessor. Stimuli delivered into the rat brain through a wireless micro-stimulator brought the rat to bear virtual cues or rewards and follow the commands to move right, left and forward in 3D obstacle environments. We designed a portable bi-directional telemetry communication system, which can deliver four-channel stimuli and receive two-channel neuronal signals in real time. It helps us in studying brain stimulation and neuronal activity recording in freely behaving rats. The firing patterns of neural units in motor cortex (M1) were detected by adjusted matched filter and classified by principal component analysis (PCA) and 3-D approximate entropy (ApEn). After synchronous analysis of the forelimb movement behavior images and neuronal signals, different statuses of neural activities during rats performing forelimb extend-flex tasks were clearly discriminated. The accuracy is 96.3%. The invasive brain-machine interfaces in rat are helpful to further explore the relationship between stimulation and responses in the neural network in free-moving status. The experiment results provide some new tools to brain-machine interfaces technologies. The system is useful for neuron behavior encoding mechanism research.