I started my internship at the Surgical Planning Lab at Brigham and Women’s Hospital with very little background knowledge. I had taken a year long computer science course, but it felt as though the material I was working with was way over my head. I had to teach myself the basics of a few new programming languages, including python, C#, and C++. The common theme shared throughout the lab was that people learned by flailing in the deep end until they could hold themselves afloat. And so, for the first few weeks, I flailed tremendously.
At the start of the summer, I was assigned to two projects. The first project had the objective of localizing bronchoscopes using neural networks. The second had a goal of using machine learning agents in order to make the process of getting an endoscope through the epiglottis an easier and less invasive one. For the first week or so, I attempted to handle both work loads, but realizing that I would be more efficient and productive working on only one project, I was allowed to drop the second project.
For the next few weeks, I worked at training the data we had and using neural networks to essentially “teach” the computer to recognize specific places in the lung alongside another summer intern from Concord Academy. Most of the early work we did was on a computer, as we were constantly training, testing, tweaking, and eventually succeeding in getting rendered depth maps that we could have for further use.
Our next step was to take the depth maps and line them up with things called point-cloud models so that we could create an accurate 3D model of the lung that was identical to the CT scan. For that portion, we were invited to a project week in July at MIT that focused specifically on the use of 3D Slicer, an extremely helpful software platform that allowed us to complete our task. We were surrounded by about thirty professionals from all over the world who were all living out careers in either software, IT, or medicine. These people offered their help to us and our project when we were feeling too stuck or unsure what to do.
I have mainly learned how to be self taught through this internship. I have relied less on other people and I have been forced to depend on myself. I will continue to complete this project up until the end of the internship, and hopefully we will have a successful product for other medical professionals to see and possibly use.
