I started this week by attending more procedures in the interventional radiology department. I did not see any new procedures. As usual, they were IVC removal or mapping of the blood vessels with the guide of a CT scan. Although the cases were interesting, I want to explore the work from other departments. For the remaining of the summer, I hope contact other physicians who specialized in neurological and cardiovascular surgery.
As for my research project, I spent most of my time reading the literature in order to understand the algorithms and the data that are used for my analysis. Dr. Kuceyeski was able to send me the dataset by friday night. To my surprise, it is one of the biggest dataset that I have ever work with. It has information about the structural and functional connectivity of the brain from dozens of subjects. In addition, it contained measures about learning, attention and memory collected over a year. Processing this type of data is truly a daunting task for me since I do not have a lot of experience with this type of statistics. However, I remain confident in implementing a machine learning algorithm to predict the outcomes of patients who suffered from traumatic brain injury. The algorithm I am currently using is known as General Regression Neural Network (GRNN), which was developed by Donald F. Specht in 1991. To my relieved, MATLAB has a version of GRNN. Unfortunately, I did not have the toolbox in my computer and Cornell does not offer a free version of MATLAB or its toolbox for the students. Therefore, my labmates spent friday afternoon helping me set up a computer with the necessary software to start my work.
As for my research project, I spent most of my time reading the literature in order to understand the algorithms and the data that are used for my analysis. Dr. Kuceyeski was able to send me the dataset by friday night. To my surprise, it is one of the biggest dataset that I have ever work with. It has information about the structural and functional connectivity of the brain from dozens of subjects. In addition, it contained measures about learning, attention and memory collected over a year. Processing this type of data is truly a daunting task for me since I do not have a lot of experience with this type of statistics. However, I remain confident in implementing a machine learning algorithm to predict the outcomes of patients who suffered from traumatic brain injury. The algorithm I am currently using is known as General Regression Neural Network (GRNN), which was developed by Donald F. Specht in 1991. To my relieved, MATLAB has a version of GRNN. Unfortunately, I did not have the toolbox in my computer and Cornell does not offer a free version of MATLAB or its toolbox for the students. Therefore, my labmates spent friday afternoon helping me set up a computer with the necessary software to start my work.
No comments:
Post a Comment