Research in brain-machine interface (BMI) technology has the potential to revolutionize how we interact with the world, offering significant advancements in healthcare, accessibility, and human-computer interaction. Wearable BMIs can provide non-invasive, continuous monitoring and control, enabling real-time applications in prosthetics, rehabilitation, and assistive technologies for individuals with disabilities. Moreover, these advancements could lead to innovations in augmented reality, gaming, and communication, enhancing everyday life and expanding the possibilities of human-machine integration. The portability and ease of use of wearable BMIs make them a practical and scalable solution for widespread adoption.
In this project, the student will be provided with a custom wireless EEG recording device developed in our lab. This EEG device streams recorded scalp EEG signals to a computer, where they are processed in a framework programmed in Python. The student is expected to propose a method to use the BMI device to control something in real time, such as robotics, electric model cars, or computer programs. To achieve this goal, the student will develop signal processing and machine learning models for real-time operation. The student is expected to collect data themselves for training the model and eventually demonstrate the system in real-time.Implantable neural interfacing technologies, especially electrical neural stimulators, are crucial for treating neurological disorders by directly modulating nerve activity, thereby restoring functions such as movement, sensation, and pain control in patients with conditions like Parkinson's disease, epilepsy, and spinal cord injuries. These technologies offer targeted and adjustable interventions, enhancing patients' quality of life and enabling precise medical treatments that were previously unattainable.
A primary concern in the design of invasive neural stimulators is safety, as residual charge can lead to tissue damage and electrode dissolution. Consequently, achieving charge balance is essential in electronic circuit design. The current mode stimulator is frequently utilized for this purpose. An ideal current mode stimulator can regulate the amount of charge irrespective of the load impedance. Such a stimulator requires a current source with high output impedance and precise current accuracy, ensuring minimal mismatches between channels.Recording brain signals from a large number of channels requires a high-speed wireless data link. Unfortunately, conventional wireless transceivers are either too slow (such as Bluetooth or Zigbee) or too power-hungry (such as Wifi). In this project, we aim to develop a low-power wireless module to achieve high-speed data transmission. We will be using a commercial Ultra-Wideband (UWB) Impulse Radio Transceiver Module, such as the SR1020 from Spark Microsystems (around 6GHz ~ 9.3GHz). https://www.sparkmicro.com/products/. By the end of this project, we aim to have a custom PCB that integrates the UWB chip with a SPI interface to an MCU.
More reading:Updated on 06/28/2024