Closed jianjun0407 closed 2 years ago
Thanks for your attention! Here are the answers to your questions: (1) In our paper we directly use the processed segments provided by the IntrA dataset, and the detailed procedure to get segments from reconstructed 3D models is shown in their GitHub repositories. I remember they're using one software named Amira, where segments can be divided manually. (2) I think directly computing the geodesic matrix on the whole cerebral vessels may require a lot of time, while calculation on segments that only contain thousands of points is quick. Besides, you can try other algorithms from modern 3D processing libraries.
Thanks for your reply and look forward to your paper.
Thank you for your work and reply, and I have the following two questions to discuss with you: (1)I have one person's cerebral vessel segmentation result, but the current network input is vessel segment, how do I divide the whole brain vessels into multiple segments? (2)As for the calculation of Geodesic matrix, I tried the method of "Geodesics in heat", but due to the large amount of calculation, the speed is still very slow. Is there a faster calculation method?