High Performance Deformable Image Registration Algorithms for Manycore Processors

• Main Authors: Shackleford, James, Kandasamy, Nagarajan, Sharp, Gregory
• Format: eBook
• Language: English
• Published: Chantilly Elsevier Science & Technology 2013; Morgan Kaufmann
• Edition: 1
• Subjects: Computer algorithms; Computer graphics; Digital techniques; Graphics processing units; Image processing; Image registration
• ISBN: 9780124077416; 0124077412; 9780124078802; 012407880X
• DOI: 10.1016/C2012-0-03039-6

High Performance Deformable Image Registration Algorithms for Manycore Processors develops highly data-parallel image registration algorithms suitable for use on modern multi-core architectures, including graphics processing units (GPUs). Focusing on deformable registration, we show how to develop d...

Table of Contents:

• 5.1 Introduction -- 5.2 Demons Algorithm for Deformable Registration -- 5.3 SIMD Version of Demons Algorithm -- 5.4 Performance Evaluation -- 5.5 Summary -- References -- 6 Plastimatch-An Open-Source Software for Radiotherapy Imaging -- 6.1 Introduction -- 6.2 Overview of Plastimatch -- 6.2.1 Automatic 3D-3D Registration -- 6.2.2 Cone-Beam CT and Digitally Reconstructed Radiographs -- 6.2.3 Interactive (Landmark-Based) Image Registration -- 6.2.4 2D-3D Registration -- 6.2.5 Automatic Feature Detection and Matching -- 6.2.6 Data Interchange -- 6.2.7 User Interface -- 6.3 Licensing -- References
• Front Cover -- High-Performance Deformable Image Registration Algorithms for Manycore Processors -- Copyright Page -- Contents -- Biographies -- 1 Introduction -- 1.1 Introduction -- 1.2 Applications of Deformable Image Registration -- 1.3 Algorithmic Approaches to Deformable Registration -- 1.4 Organization of Chapters -- References -- 2 Unimodal B-Spline Registration -- 2.1 Introduction -- 2.2 Overview of B-Spline Registration -- 2.2.1 Using B-Splines to Represent the Deformation Field -- 2.2.2 Computing the Cost Function -- 2.2.3 Optimizing the B-Spline Coefficients -- 2.3 B-Spline Registration on the GPU -- 2.3.1 Software Organization -- 2.3.2 Calculating the Cost Function and ∂C/∂ν&amp -- #8594 -- -- 2.3.3 Calculating the Cost Function Gradient ∂C/∂P -- 2.4 Performance Evaluation -- 2.4.1 Registration Quality -- 2.4.2 Sensitivity to Volume Size -- 2.4.3 Sensitivity to Control Point Spacing -- 2.5 Summary -- References -- 3 Multimodal B-Spline Registration -- 3.1 Introduction -- 3.2 Using B-Splines to Represent the Deformation Field -- 3.3 MI as A Cost Function -- 3.4 Efficient Computation of MI -- 3.4.1 Constructing Histograms for the Static and Moving Images -- 3.4.2 Constructing the Joint Histogram -- 3.4.3 Evaluating the Cost Function -- 3.4.4 Optimizing the B-Spline Coefficients -- 3.5 Performance Evaluation -- 3.5.1 Registration Quality -- 3.5.2 Sensitivity to Control-Point Spacing -- 3.6 Related Work -- 3.7 Summary -- References -- 4 Analytic Vector Field Regularization for B-spline Parameterized Methods -- 4.1 Introduction -- 4.2 Theory and Mathematical Formalism -- 4.3 Algorithmic Implementation -- 4.4 Performance Evaluation -- 4.4.1 Registration Quality -- 4.4.2 Sensitivity to Volume Size -- 4.4.3 Sensitivity to Control-Point Spacing -- 4.5 Summary -- References -- 5 Deformable Registration Using Optical-Flow Methods