Week 4: If CAT-scan do it, so can I!

During this week, I got more involved in the research and obtained preliminary results for the research project. First, I segmented the prostate, and its regions, the transitional and peripheral zones in MRI scans. Then, I used the PMOD software, which calculated the different types of texture analyses in the segmented areas. I imported the data into MATLAB and plotted the MRI texture for all patients against the PET texture to see whether there would be a linear or other types of correlation between the MRI and PET scans. Because we had only three useable patient cases, we were unable to see a correlation, so we need more patient data to draw a conclusion.

I also compared the MRI pixel intensities and the standardized uptake value (SUV) from the PET images. The SUV is proportional to the amount of radiation given off by the germanium-68 in the body. Before PET imaging, the patient is injected with germanium-68 bonded to prostate specific membrane antigen (PSMA) which has a high specific binding to the tumorous lesions in prostate cancer patients. Because the SUV is related to the PET scan pixel intensity, we compared the MRI pixel intensity and the PET pixel intensity, a.k.a the SUV. However, this set of analyses also only has three samples, and has not been analyzed yet. One of main challenges of this project was trying to organize the large sets of data to make them ready for analysis. I realized that especially with large data, it is important to visualize the data placement, so that you can easily find the data. However, deciding how to group the data so that it would be easy to locate was more difficult than I expected. During this process, I became more familiar with MATLAB programming and learned new ways of displaying information.

I also shadowed Dr. Martin Prince and saw how MRI images are taken.  I learned that often it is difficult to obtain patient consent about using their images for research purposes, and that is more important to make the patient feel comfortable about using their information for research or not, rather than obtaining data for a study. I also saw a patient with osteosarcoma on their limb and was surprised to hear Dr. Prince mention how the skin near the tumor was hotter than the skin far away. This effect might be caused due to the higher metabolic activity that occurs around a tumor. The tumor also seemed to have expanded to the point of pressing on the vascular walls and constricting the veins. This prevented blood near the tumor from leaving that region and caused swelling in that region. A type of imaging, called thermal imaging, looks at the heat distribution is tissue and is able to identify a tumor based on the higher variation in heat distribution. By next week, I hope to have more patient scans to make conclusions about my research project.

The highlight of my week, and probably the whole summer was getting to see the Lion King on Broadway! Although our seats were in the very last row, the show was an amazing and a mind-blowing experience. Thanks to Jacob who graciously invited me to see the show!