Well, I am back in Houston now and I got a chance to start implementing some of the stuff I have been mentioning in my past few posts. Essentially, rather than running the algorithm one time with a single set of parameters, I am running the same algorithm N time. The initial results are promising as the algorithm is able to successfully track essentially all of the arteries for which a meaningful result can be expected.
The next major challenge, now that I have an algorithm which can successfully perform the task, is to find ways to decrease the run time. Each iteration requires about 200 msec to run, but I think I can get it down to about half of that.
The other major issue that I have to contend with next is determining how to proceed from here. The algorithm has reached a level of robustness in the vascular and bone segmentation that it is time to complete the job with the sub-segmentation and move forward to the individual organ segmentation. Before I move on to the next stage of the algorithm which could take a significant amount of time, I think I should publish the results of what I have completed so far. I don't know of any algorithms that currently exist which can track the kind of low contrast vessels that this algorithm is successful with.