Bryan Hsieh

EEG Datalogger:

• Developed a miniature ( < 20mm by 17mm ), low power ( < 3mW ) neural logger to research sleep circuits in neuroscience.

• Proposed and implemented a highly accurate, unsupervised online/offline sleep stage scoring algorithm based on Gaussian Mixture Model and z-score normalization.
• Designed a custom multi activity Android BLE application for real time data verification and device control.
• Designed a custom GUI application based on the MVC framework using Qt creator for data acquisition and device configuration.
• Produced a BLE firmware for an ARM based SoC and highly optimized it for power and speed via the usage of DMA, data buffers and hardware interrupts.

• Designed a custom optogenetic coupler for optical fibers for closed loop stimulation in SolidWorks

Dementia Research:

• Wrote an automatic script to analyse, sleep score and collate large datasets of EEG and photometry recordings in MATLAB for Dementia research.

Rodent Tracking :

• Designed a system to automatically deliver video stimuli upon detection of movements using Raspberry Pie, OpenCV2 and Python.

• Implemented a DNA alignment algorithm based on Burrows-Wheeler

transform for matching DNA reads to a known genome on multiple platforms.

• Investigated OpenCL as an alternative to HDL for programming FPGA's using Altera's OpennCL SDK and Stratix V board

• Compared the performance of the algorithm across different platforms and also against other existing algorithms such as Soap2 and Bowtie.