SUMMARY:

​

Over the winter break, I attempted to work through the second and third phases of my projects and ran into a roadblock regarding data and novelty. I reached out to a Harvard Researcher and met with them to briefly discuss this issue, as they had worked on the dataset a few years prior in a different setting. I made some significant changes to my project based on the suggestion to remove GANS from my project. I applied for the Kaiser Permanente Hospital Readmission Dataset independently and removed GANS from the second step of my methodology, replacing it with the less computationally intensive and more reasonable to execute Semi-Supervised Learning and Saliency Mapping, both of which also had better physician supplementary applications.

​

12/16 5:00 PM: While reading through the literature of the Framingham Heart Study, I found that the dataset of CT scans was limited to later cohorts than the one I had access to. Additionally, the CT Scans of this CT cohort were relatively low in number. 

​

12/18 3:23 PM: Applied for the Multiethnic Study of Atherosclerosis Dataset after being directed to it by Dr. Borek Foldyna nearing the end of a research meeting

​

12/20 1:54 PM: Request status updated to "Pending Requestor Documentation" by Rebecca Myers on the National Heart Lung and Blood Institute (NHLBI) Website. 

​

12/21 1:00 PM: Met in person with a Harvard Researcher where I presented my project in its previous form "Evaluating the Significance of CHD Risk Factors and Generating Images of CT Scans using ML"

​

12/21 2:00 PM: The Harvard Researcher suggested that I drop GANS from my project because of its complexity and computing power requirements. They had used GANS in image generation in a Parkinsons Related Project a few years prior

​

12/22 10:03 PM: I Began applying for the Kaiser Permanente Hospital Readmission Prediction Dataset

​

12/23 3:14 PM: MESA Study Request status updated to "Closed - Request Requirements" on NHLBI

​

​

​

 

​