Virtual planning of coronary bypass surgery generated the need to perform real-time adaptation of lengths estimations of implanted grafts. Specifically, the adaptation should fit the actual inflated heart dimensions. Furthermore, due to time constrains, existing elastic 3D registration during surgery are prohibited. In order to solve this problem, we created the H2 algorithm as an efficient mean for estimating grafts' lengths for asymmetric 3D heart volume. The H2 algorithm defines asymmetric digitized chain-code for the implanted grafts represented as curves. It uses 43 length primitives which can be adapted in real-time to any asymmetric deformation of the underlying 3D grid. This algorithm does not require the usage of registration methods or intense computation effort. It eliminates the need to apply spatial transformations by using specific chain codes. This elaborated chain code is twice faster than existing ones.
[1]
Benoit M. Dawant,et al.
Non-rigid registration of medical images: purpose and methods, a short survey
,
2002,
Proceedings IEEE International Symposium on Biomedical Imaging.
[2]
Patrick Clarysse,et al.
A review of cardiac image registration methods
,
2002,
IEEE Transactions on Medical Imaging.
[3]
S. C. Lee,et al.
The C-chain code-a new method for coding 3D curves
,
1991,
[1991] Conference Record of the Twenty-Fifth Asilomar Conference on Signals, Systems & Computers.
[4]
Wan Rui,et al.
An overview of medical image registration
,
2003,
Proceedings Fifth International Conference on Computational Intelligence and Multimedia Applications. ICCIMA 2003.