Biomedical Image Analysis

Segmentation With Local Adaptive Thresholds and Related Methods (Biomedical Image Analysis)

Thresholding is one of the most common segmentation methods. The underlying assumption is that the feature (foreground) pixel brightness differs from the background pixel brightness. This behavior is usually evidenced by a bi- or multimodal histogram, where one peak represents the background and one (or more) peaks represent the features (see Section 2.4). If the […]

Biomedical Examples (Biomedical Image Analysis)

Adaptive methods have been under development for many years and are finding their way into biomedical image analysis. In most cases, adaptive filters constitute an intermediate step in the image processing chain. For this reason, adaptive filtering is found most frequently as a tool rather than as the research objective of a study. With the […]

Deformable Models and Active Contours (Biomedical Image Analysis)

In simple terms, two-dimensional deformable models (two-dimensional closed-contour deformable models are also called snakes) can be thought of as the numerical modeling of rubber bands subjected to image-dependent external forces. Deformable models are used in image processing to delineate and segment features interactively. The most interesting property of deformable models is their closed shape, even […]

Two-Dimensional Active Contours (Snakes) Part 1 (Biomedical Image Analysis)

Greedy Snake Algorithm Probably the most widely used implementation of a snake is the greedy snake algorithm introduced by Williams and Shah.34 Some investigators noted stability problems with the original numerical implementation by Kass et al.16 and presented alternative approaches to improve numerical stability and to reduce the tendency of snake vertices to cluster around […]

Two-Dimensional Active Contours (Snakes) Part 2 (Biomedical Image Analysis)

Gradient Vector Flow In the context of the greedy snake algorithm, multilevel blurring was mentioned as one method to increase the capture range of the snake. Another powerful method to increase the capture range and to provide an external force field that stabilizes the behavior of the snake is to diffuse the edge image. Xu […]

Three-Dimensional Active Contours (Biomedical Image Analysis)

The snake, a one-dimensional curve that encloses a two-dimensional image feature, was discussed in the previous sections. The governing equations can easily be extended to three dimensions. For example, the balance of forces described in Equation (6.7) would be discretized through finite differences in the x, y, and z directions. The external force (image force) […]

Live-Wire Techniques (Biomedical Image Analysis)

The live wire is a highly interactive segmentation technique, aimed at aiding the user delineate an object boundary. The user selects a start point on the boundary. As the user hovers over the image with the mouse, the live wire tries to connect both points with an energy-minimizing path. The live wire tends to snap […]

Biomedical Examples (Biomedical Image Analysis)

The primary use of active contours in biomedical imaging is user-guided segmentation. Active contour models help improving accuracy and repeatability of the segmentation results.2 In addition, user-guided segmentation techniques such as snakes and active contours speed up the segmentation process, since the user specifies only a few points instead of tracing the entire contour. After […]

The Hough Transform (Biomedical Image Analysis)

The Hough transform is a tool to detect and quantify shape primitives in images, particularly in the presence of noise. The Hough transform is a robust tool to extract features (such as straight edges, circles, or ellipses, but also primitives defined by polygons) from images and describe them parametrically. The Hough transform is generally used […]

Detecting Lines and Edges With The Hough Transform (Biomedical Image Analysis)

To use the Hough transform to extract lines, the equation of a line must be rewritten to prevent undefined values that occur in Equation (7.2) for vertical lines. Instead of using slope and intercept, a line can be described by its distance to the origin, p, and its angle with the x-axis, 0 (Figure 7.2). […]