Engineering and Science

As an independent final project for my Bachelors degree in Electrical Engineering, I wrote a software implementation of an artificial neural network algorithm to distinguish different shapes in images.  To ensure that I was learning engineering and not just neural networks, the faculty required that I wire wrap a photodiode-based scanner including the analog-to-digital and serial communications circuitry to use as the image input device.

As an independent final project for my Masters degree in Electrical Engineering, I designed a hardware implementation of an artificial neural network learning rule that I invented.  I integrated the learning rule into my coursework on neuromorphic VLSI chip design even though an application-specific circuit model was probably not the most efficient means of studying its capabilities.

For my second graduate degree, my intent was to continue to study my learning rule, but this time within a school of science rather than engineering.  I would focus on Computational Neuroscience which models neuronal network circuits using software simulations.

I was distracted from my original plan, however, by an increasing interest in neuroprosthetics, specifically wireless neural interfaces for controlling robotic arms.  After I finished my Masters in Applied Cognition and Neuroscience, I submitted a related grant proposal as part of my doctoral studies toward a degree in Cognition and Neuroscience.  I was told by faculty that my proposal was rejected because the literature review section contained too many references to engineering papers and not enough to neuroscience.  Upon review, I noted that the disallowed citations were from journals of neural engineering and neural surgery.

Since then, I have continued my part-time doctoral studies by exploring Computational Neuroethology (CNE), the study of how simulated neuronal network circuits create behaviors when embodied in robotic or virtual reality environments.  My previous studies of biologically-plausible learning algorithms and body-to-machine neural interfaces fit neatly within CNE.  Furthermore, practitioners of CNE avow that it is a scientific discipline, not engineering, and therefore I assume it is valid subject matter for a thesis defended within a School of Behavioral and Brain Science.  There is some risk, however, that those who consider Computer Science not to be a true science might object to CNE on the same grounds.

Fortunately the University of Texas at Dallas (UTD) Erik Johnnson School of Engineering and Computer Science has just started a new graduate degree program in Biomedical Engineering (BME).  Neural Engineering is classified as a subdiscipline of BME.  I am now taking the introductory course "Anatomy and Human Physiology for Engineers" and I am enjoying it.  I am considering pursuing a third Masters within an academic department that embraces both engineering and science.