NRG Lab 2008: (Top row) Dr. Sanchez, Sankaraleengam Alagapan, Babak Mahmoudi, Aysegul Gunduz, (Bottom Row) Erin Patrick, Jack DiGiovanna, Scott Amerman
Ph. D. Students
Aysegul Gunduz
My goal is to build Brain Machine Interfaces (BMI) using electrocorticogram recordings (ECoG) from human patients. This is a big step towards non-invasive BMIs compared to BMIs utilizing single unit activities recorded by invasive microelectrode arrays. With efficient online channel selection algorithms, I aim to build models that not only work offline on recorded data, but also in closed loop interactions. Moreover, I want to test new experimental paradigms by immersing patients into (virtual) environments with acting physical forces (e.g. gravitational), and achieve adaptability to one’s changing surroundings.
John DiGiovanna
Jack earned a M.E. in biomedical engineering from the University of Florida, Gainesville, in 2007 and is working to complete a PhD. He earned a B.S. in electrical engineering (minor in bioengineering) from Penn State University, University Park, in 2002. He joined the CNEL lab in 2004 and NRG lab in 2006. Jack’s research area is reinforcement learning based brain-machine interface (BMI). He also has research interests in motor control systems. In 2007 he received an NSF International Research in Engineering and Education grant and worked with the Sensory Motor Control Group at Cambridge University and the Advanced Robotics Technologies and Systems lab at Scuola Superiore Sant’Anna. He was a founding officer in the Gainesville EMBS chapter. He holds one patent in neuroprosthetic design and is the author of 7 peer reviewed papers.
Erin Patrick
My role in the brain-machine interface (BMI) project is the design and fabrication of flexible substrate microelectrodes. Our goal is to understand the issues with chronic recording and design the microelectrodes to promote a long recording lifetime. We currently have a design that incorporates a flexible polyimde cable as the substrate for metal probes. It has been tested successfully in an animal model. Future research focuses on materials used as the electrode as well as surface modification techniques to promote a more compatible interface with brain tissue.
Babak Mahmoudi
As a Biomedical Engineer, my interest in brain machine interface research is due to the interaction of two intelligent systems in this area; Brain and Machine Intelligence. In the Neuroprosthetics Research Group I am working on developing a closed loop BMI, based on an information processing model of the brain for hand movement and maximization of mutual information. In this paradigm, the processing model of brain combines the environmental information with its internal states in a data fusion process.
Graduated Ph. D. Students
Yiwen Wang
I am PhD student in the group of Brain-Machine Interfaces (BMI). My work is to design, improve, and apply the signal-processing algorithm for the mapping between the kinematics and the neural spike trains. There are a lot of challenges: the better understanding of the underlying dynamics of the neural receptive field, a nice model for the nonstationary neural behaviors, the association between neurons, and finally the better adaptive filtering tool to make the better estimation. I am now interested in the inhomogeneous Poisson model of spike trains, and Monte Carlo adaptive filtering algorithm to infer biological signals from point processes. It may take years to struggle with them all ?That'swhy I am pursuing a PhD here.
Christy Rogers
I was a member of the Brain-Machine Interface (BMI) project. My role was to design and implement analog VLSI hardware to perform spike feature extraction and/or detection in extracellular neural recordings. I was awarded a National Science Foundation Fellowship to help support my studies and research.
Yuan Li
I was a member of the Brain-Machine Interface (BMI) project. My work was concentrated on analog hardware design. Specifically, it was to design and implement biphasic integrate-and-fire(IF) neuron to encode the analog neural signal into time domain spike train. I'm was also working on multi-channel readout circuitry design.
