Table 2 Tagging analysis techniques

Active contour

Uses spline curves that are fitted to the tag lines using multiple constraints.

Intuitive approach; parametric continuity; local control of the curve shape.

Long processing time; sensitive to weights of different constraint forces.

93, 96, 97, 101, 102.

Optical flow

Tracks tag lines intersections based on tagging contrast.

Possibility for automatic processing; reduced processing time.

Sensitive to image quality, especially tagging contrast.

103-105, 107-110.

Template matching

Cross-correlates a pre-defined tagging pattern with the resulting images.

Reduced processing time.

Pre-defined assumptions must be met.

95.

Sinusoidal analysis

Data are analyzed into different frequency components.

Decreased sensitivity to noise; high accuracy.

Complicated data analysis.

111-113.

Volumetric modeling

Analyzes a stack of parallel tagged images.

3-D tagging analysis; more automatic processing.

Long processing time.

134-139.

Finite-element modeling

Creates model tags, which define the tag lines in the images.

3-D tagging analysis; reduced processing time.

Measurements are not directly related to clinical understanding.

140-142.

Statistical modeling

Uses statistical methods for estimating tag lines deformation.

3-D capability; more intuitive and understandable parameters.

Predefined assumptions; complicated processing.

141, 143.

3-D active contour modeling

Uses 3-D spline curves that are fitted to tag lines from a set of parallel images.

3-D capability; high resolution; parametric continuity.

Long processing time.

144-148.