Wednesday, August 1, 2007

Calcium Score: Part 2

Okay… I think I promised you that I would upload something on a new GE software, which I was supposed to use this past week to measure calcium score. Apparently, there are some legal issues in distributing the software so I don’t think I’ll be using it before we all leave for Ithaca. In a perfect world, I would have compared the calcium score measurements made by Smart Score (discussed last time) to those made by VCAR (the new GE software). Instead, I will only have the Smart Score measurements done and my mentor, Dr. Min, will finish the VCAR measurements whenever he gets the software. I think I will still get my name on a paper since he plans to send out a manuscript highlighting the results of the comparison.

Maybe it is better in a certain way that I will not be collecting scores using VCAR. According to my mentor, although the software is more advanced than Smart Score, it is just as tedious to use. If you recall, the Smart Score software forced one to highlight the calcium in transverse CTA slices. It, thereby, was able to deduce a calcium score based on just volume (volumetric score) or area and average Hounsfield value (agatson number) for the overall calcium in the arteries. VCAR calculates the score in a similar fashion, but it does not need the user to highlight the calcium. Instead, it is designed to automatically segment the calcium in the arteries. However, before the algorithm is capable of doing this, one must click along the centerline of an artery so that the algorithm can basically fit a curve to the selected points and subsequently subtract it from the image to straighten out the artery. Then it presents the artery of interest in a longitudinal format so that one is better able to view the calcium deposits (figure). I’m not certain why this step is necessary. Nevertheless, the software seems very user friendly so I think it would have been fun to work with.

No comments: