|Email:||sip2109 (at) barnard.edu|
|Home Institution:||Barnard College, Columbia University|
|Project:||vasAR visualizing catheters|
My project is focused on visualizing Catheters in real time in 3D models in Unity. Currently, vasAR, which I am working on, is a system to help physicians visualize 3D models during non-invasive procedures. My goal is to add a live 3D catheter to these procedures.
I am doing this by converting existing Matlab code into Unity C#, doing the conversion by hand and adding functions.
I got the computer working with the hololens, which was refusing to connect to the computer via USB. I also learned how to do basic voice commands for the hololens and inserted a few of my own.
I learned more about how complex voice commands for the hololens worked, and was able to let people insert arbitrary numbers using an XML grammar. I also presented my project to the REU.
I met with Shawn and Shirin, and we discussed the major focus of my project. I was given a paper, linked in recources, which I read and annotated and figured out how to use. I then looked at the code they provided, in Matlab, and verified that it worked. I started the process of importing it into Unity with the calcZ file.
I put more of a focus into making the code more readable and future proof, making sure everything was clear, retroactively and moving forward. I established a standard of how to store images for manipulation. I continued with the conversion, with the establishment of the subtract, binarize, and gaussian filter functions.
More of the same. The standards I was working with were mostly set and the bases for things mostly there. This week the focus was on the area filter, gradient, and logical functions. There was also other code being converted, but those were the big and important things.
More of the same. The standards I was working with were mostly set and the bases for things mostly there. This week the focus was on the morphological, structuring, and dilation functions. I figured if I was already reinventing the wheel I might as well do my own slightly custom sort and find functions. As with last time, there was also other code being converted, but those were the big and important things.
I talked with Shanw again about progress. I was able to 3Dify the points. A main focus of my work was region masking and fitting functions.