Canny edge-based deformable image registration

Abstract

This work focuses on developing a 2D Canny edge-based deformable image registration (Canny DIR) algorithm to register in vivo white light images taken at various time points. This method uses a sparse interpolation deformation algorithm to sparsely register regions of the image with strong edge information. A stability criterion is enforced which removes regions of edges that do not deform in a smooth uniform manner. Using a synthetic mouse surface ground truth model, the accuracy of the Canny DIR algorithm was evaluated under axial rotation in the presence of deformation. The accuracy was also tested using fluorescent dye injections, which were then used for gamma analysis to establish a second ground truth. The results indicate that the Canny DIR algorithm performs better than rigid registration, intensity corrected Demons, and Distinctive Features for all evaluation matrices and ground truth scenarios. In conclusion Canny DIR performs well in the presence of the unique lighting and shading variations associated with white-light-based image registration.

A continuous arc delivery optimization algorithm for CyberKnife m6

“…To obtain the minimum distance required for the robot to travel to every , obtained from final set of beam geometries in step one, the problem is cast into a constrained 3D genetic optimization algorithm . The genetic algorithm solves the traveling salesmen problem by selecting a minimum path from a population of paths.…”

“…To obtain the minimum distance required for the robot to travel to every K final , obtained from final set of beam geometries in step one, the problem is cast into a constrained 3D genetic optimization algorithm. [15][16][17] The genetic algorithm solves the traveling salesmen problem by selecting a minimum path from a population of paths. A new generation (gen) of populations (pop) is created based on the previous population's minimum path, until a globally minimum path k final is determined.…”

Computed Tomography Imaging Based on Edge Detection Algorithm in Diagnosis and Rehabilitation Nursing of Stroke Patients with Motor Dysfunction

“…e Canny edge detection operator is different from other operators such as Sobel in which it firstly smooths the image and then finds the derivative, and the derivative can detect the edge. Some research experts have also used the Canny edge detection algorithm to improve the edge blur caused by the compressed sensing algorithm in CT image reconstruction, and the obtained effect is good [15,16]. However, it is found that the Gaussian filter is not ideal for the treatment of scattered salt and pepper noise, and the dual threshold of the Canny operator is mainly set by manual experience, which is poor in generality and compatibility in practical applications.…”

A convolutional neural network algorithm for automatic segmentation of head and neck organs at risk using deep lifelong learning

“…With atlas‐based autosegmentation, a new set of contours based on a library of previous patient contours is generated, using a patient selected from an atlas to deform previous contours on to new anatomy . However, this approach relies on deformable image registration, which remains a potentially inaccurate process . Other investigators have implemented a variety of machine learning algorithms, such as principal component analysis (PCA) or random forest (RF) methods, to predict contours for OAR .…”

“…[3][4][5] However, this approach relies on deformable image registration, which remains a potentially inaccurate process. [6][7][8] Other investigators have implemented a variety of machine learning algorithms, such as principal component analysis (PCA) or random forest (RF) methods, to predict contours for OAR. [9][10][11] However, the highly dimensional nature of auto-contouring is challenging for linear models such as PCA, while RF-based methods rely on structured datasets, which require reducing a highly informational three-dimensional (3D) image into a sparse representation of features.…”