Q&A: Guy Hotson
Guy Hotson is a PhD student from Palo Alto, CA. He received his undergraduate degree in computer engineering, with a minor in bioengineering, from Santa Clara University. You can learn more about his research into a “mind-controlled prosthetic arm” in the Baltimore Sun.
How did you get interested in electrical and computer engineering? What interests you the most about the field?
I got interested in computer engineering during my freshman year of undergrad. I took an introductory class on the broad fields of engineering, and it amazed me how electrical/computer engineering skills could be used to solve real-world problems just by sitting down at a computer. Projects such as helping build a database for the California Rescue Dog Association made me fall in love with the major. I later decided to come to Johns Hopkins for my PhD, where ties with the prestigious medical institute enable projects that can immensely benefit patients in need.
Your current project—the “mind-controlled prosthetic arm”—is getting a lot of attention. How did this project come about?
The efforts to obtain neural control of an upper limb prosthetic began with the previous generation of PhD students in Dr. Thakor and Dr. Crone’s lab working in collaboration with the JHU Applied Physics Lab. They laid down the groundwork that made this work possible during the Revolutionizing Prosthetics 2009 effort.
The decoding of individual finger movements in a human subject was the product of preparation and good fortune. The location/type of neural implants are dictated by clinical need, so we have to be prepared for any possible coverage. So the focus on individual finger decoding in particular didn’t begin until we got amazing coverage of the sensorimotor hand areas with cutting edge high-density electrode coverage. This happened about a year ago, at the beginning of 2015.
The biggest obstacle has been the limited amount of information able to be captured by the neural signals. The placement of electrodes is done entirely according to clinical need, meaning we typically get coverage that isn’t ideal for our work. And even when the electrodes are placed over areas needed for our studies, only so much information can be extracted from the signals. This means great care needs to be taken when extracting features from the neural signals, and we need a thorough understanding of how these features of the neural signals relate to the subjects’ movements.
What impact do you hope to make through your research?
I hope my PhD research helps progress neurally controlled prosthetics another step closer to restoring autonomy to paralyzed patients. More broadly, my career goal is to use machine learning and intelligent robotics to help people. While most of the tangible benefits of this line of research won’t be seen for years, the end results will vastly improve countless lives.