Implantable medical devices such as brain machine interfaces (BMI), deep brain stimulators (DBS), vision prosthesis and cochlea implants use electric pulses to stimulate the central nervous system. These electronic devices have restored hearing to the deaf, functional sight to the profoundly blind, alleviate symptoms of drug resistant depression and epilepsy, and allowing those with locked-in syndrome â€“ the inability to move and speak due to brainstem injury or ALS â€“ to communicate and move again.
The parameter space for these stimulation pulse trains are enormous, involving for example, waveform shape, amplitude, duration, inter-pulse interval and phase. The pulses used in existing devices are based on educated guesstimation, and labour-intensive trial-and-error experiments. The combinatoric design space cannot be adequately nor efficiently addressed by such traditional, low-throughput technique. In this project you will help develop automated, high-throughput methods for finding effective stimulation waveforms.
Brain machine interface, machine learning
You will work with a team of leading neural engineering researchers at the Graduate School of Biomedical Engineering, using cutting edge techniques, such as machine learning, high performance computing clusters and cloud computing. Depending on progress, we may even test the results in biological experiments.
In this project you will use and help advance the techniques we are developing for effective and efficient stimulation waveform design.
You will gain first-hand experience in