The NURBS book
Deformable B-solids and implicit snakes for 3D localization and tracking of SPAMM MRI data
Computer Vision and Image Understanding
Shape and Nonrigid Motion Estimation Through Physics-Based Synthesis
IEEE Transactions on Pattern Analysis and Machine Intelligence
| Hi-index | 0.00 |
The paper demonstrates accurate, anatomically correct 3D visualization of left and right ventricles of the human heart from tagged magnetic resonance imaging data. Tagged MRI reveals 3D structural and motion information that is incorporated into the visualized models. We use Non-Uniform Rational B-Splines (NURBS) for our purpose, that allow non-uniform knot specification and weight specification of control points, an advantage over traditional uniform B-Splines. NURBS algorithms are numerically stable and fast. The myocardial models serve the purpose of (i) accurate visualization of tag planes and myocardial walls, (ii)checking tag tracking accuracy and registration errors, (iii)providing framework for 4D visualization and motion analysis, and (iv)obtaining accurate, memory efficient and compactly supported models of the heart in anatomical dimensions. We obtain excellent results on normal as well as pathological patient datasets.