A High Data Rate Wireless Brain Computer Interface

Book excerpt: Brain computer interface (BCI) taps into brain state and provides an associated quantitative assessment. It has potential for the real-time assessment of a person's cognitive state. BCI is a study to analyze characteristics of brain (Electrophysiology). Electrophysiology can be categorized into Invasive and Non-invasive methods. Electroencephalography (EEG) produced spontaneously during the process of thinking with frequency bands alpha waves, beta waves, theta waves and delta waves is one of the Non- Invasive methods. Electrocorticography (ECoG) is one of the invasive methods which records mu, beta and gamma frequency bands. In single unit recording, single or multi-neuron spiking and local field potential (LFP’s) are recorded. Invasive methods require surgery to implant electrodes inside the skull. A wireless interface is required to transfer brain data from BCI circuitry to host interface. Wireless transmission module eliminates the inconvenience of wiring and reduce risk of contamination. The wireless link implemented in this study has ability to switch among classic Bluetooth protocol, Bluetooth smart protocol and proprietary protocol depends on desired throughput, Power consumption and transmission distance (RSSI). Invasive methods like single unit recording requires processing high frequency band spectrum and optimum data throughput to transmit spike data. This study proposes a high speed sampling algorithm to increase sampling rate and a proprietary wireless protocol to transmit high sample data and reduce packet error rate for physiological sensing. A low-power bio potential amplifier in BCI circuitry converts analog brain data to digital format. High sampling rate can be achieved by multi-threading software call to receive data from sense electronics. The proposed sampling algorithm proposed increases sampling frequency from 800 Samples/sec to 10k for samples 8 channels. Wireless interface which allows throughput and power optimization can be achieved by implementing transmit mode with dynamic connection interval. Packet error rate is vital parameter for data integrity and can be optimized by implementing dynamic channel shifting mechanism in the presence of noisy channel. Dynamic transmit buffer length allocation algorithm should be implemented for better utilization of channel.