It’s been a while longer since I last wrote than I though thought it had been. Not all that much has really happened in the past few weeks, aside from two of three submitted papers getting accepted (I suppose this means I ought to start my Curriculum Vitae); this means I’ll likely be going to CVPR in mid-June.
In the past week or two, I’ve also laid the groundwork for an opportunity to collaborate with a neuroscientist at the Anschutz Medical Campus of the University of Colorado’s Denver branch thanks to the research that Dr. Rory Lewis of the UCCS Computer Science department is doing. Rory and Dr. Andrew White (the neuroscientist) are attempting to use Machine Learning techniques to detect and predict when a seizure will occur in epileptic patients by studying electroencephalogram (EEG) data. My piece of the puzzle will be (thanks to my background in computer vision and neuroscience) to do a similar thing from the visual display of the EEG data that Hospital technicians see. The motivation goes something like this:
Hospital technicians with no specific training in how to recognize neural activity preceding epileptic seizures from visual EEG readouts are nonetheless able to do so after several years experience, yet they are unable to explain why. This is a psychological phenomenon that is not restricted only to the medical field (Jonah Lehrer gives a similar example of an artillery captain on a British ship during the first Gulf War and his reaction to a particular blip on a radar screen) and is still poorly understood. However, in this particular case, it is not necessary to understand what the technicians are seeing, but to reproduce the results using a computer. If that goal is possible, then pattern analysis techniques can be applied to the resulting system, which are more easily obtained and analyzed. In theory, these results can then be compared with analogues from actual technicians.
This will technically be as an independent study over the summer, but, since I’ll actually be busy with vacations for half of the summer semester, it’ll probably end up being more toward the end of the spring semester and I wouldn’t be surprised if it continued into the fall semester. Obviously I have deliverables that have to be done by the end of the summer semester for the grade, but that doesn’t preclude me from continuing to do research in this area.
I also took a nice spill off my bike last weekend that is healing up nicely. It consisted mostly of road rash on my upper right arm, a nice set of gashes below the right elbow, a raw spot just below my right knee, and a chunk of skin from the heel of my right hand. Aside from that, there were a few scratches on my right side and on my leg where I landed on the contents of my pocket, as well as a few scrapes on the knuckles of my left hand.