At the beginning of this immersion program I had stated that I would be documenting my progress as a novice researcher at the clinical level. Throughout my time at HSS I have come under the mentorship of Dr. Hollis Potter, Dr. Mathew Koff (a Ph.D. kind of doctor), and Drs. Russell Meyer and Tate Greditzer, which has allowed me to see how expertise and experience establish differences in the research experience for all those in a laboratory. After I was initially delegated to looking at mapping of cartilage between the trapezium and metacarpal of the thumb joint, I was finally chosen to lead the new territory of expanding the application of MDCT towards the clinical level, which has meant that I have looked so far over hundreds of images of sheep femoral condyles in order to establish whether high quality imaging could be accomplished without the need for hazardous radiation doses, the key obstacle in the way to applying MDCT scanning to humans.
After quite a few revisions and with even more pictures coming my way, this week was actually devoted to more developmental, engineering work as I shifted from doing a job that any other competent intern could do towards work that actually did required to go into uncharted territory. My task was to develop a software program that could isolate graphical information about the trabecular microarchitecture highlighted in the femur pictures in a relatively noiseless format, translate it to a numerical format and then analyze it for key information such as number of trabeculae, trabecular depth, and trabecular width. The program would require certain aspects of image filtering, data analysis and plenty of GUI tools to acquire information about the picture. While I had seen all of these back in undergrad, I found myself spending most of the first day just basically going over how to read a picture in MATLAB, followed by the countless of operations that one can do with image data. Who knew that there are so many ways to "transform" image data?
For most of the second and third day I spent back and forth between talking with Drs. Russ and Tate, taking more pictures of even more femoral condyles, and implementing the most basic filtering algorithm to black-and-white pictures (the contrast between the graph line and the the dark background makes filtering rather easy), coming closer to finishing the program. Yesterday I worked on the data analysis part of the program, which meant a complex combination of for and while loops which essentially would not work until they did. Towards the end, I unceremoniously ended my week by trying to understand how to add a title to a simple histogram, a function that behaves just like any other plot it seems. For the most part, I believe the program is mostly done so I have since arranged a meeting with Dr. Potter to discuss new ways in which to tweak the program (such as defining what length constitutes trabeculae and what constitutes noise). Perhaps a FFT may help in the analysis...
The past weekend was devoted to another aspect of life: food. I had the pleasant experience of several serendipitous gastronomical discoveries in the downtown neighborhood of Soho, just south of NYU. The area had, unbeknownst to me, a Spanish delicatessen that offered a variety of traditional tapas and pintxos, a Basque take on the tapa. The food was authentic and rather delicious, plus the delicatessen also boasted a store that specialized on Spanish wines and sherries. Just west of this place was a small brunch place, Cocotte, which specialized on traditional French fare. From the scrambled eggs with mushrooms to a clafoutis, brunch had never been so enticing. Other highlights included a visit to a small bookshop, a wine tasting, and watching the 4th of July fireworks by the Manhattan Bridge. This weekend will be devoted to exploring Brooklyn once again, as well as doing downtown shopping for gifts and souvenirs to bring back home.
Restaurant of the week (tie):
Despaña Fine Foods and Tapas Cafe
Cocotte
No comments:
Post a Comment