image_similarity: Compute Image Similarity Metrics

In muschellij2/extrantsr: Extra Functions to Build on the 'ANTsR' Package

Compute Image Similarity Metrics

Description

Compute Image Similarity Metrics

Usage

image_similarity(image1, image2, image1_mask = NULL, image2_mask = NULL, ...)

Arguments

image1	Object of class nifti or character filename. Passed to the fixed image in imageSimilarity
image2	Object of class nifti or character filename Passed to the moving image in imageSimilarity
image1_mask	Mask for first image (optional)
image2_mask	Mask for second image (optional)
...	Additional arguments to pass to imageSimilarity

Value

A numeric value

Note

Not all metrics are symmetric.

Examples

library(ANTsR)
library(extrantsr)
x =  getANTsRData( 'r16' )
y =  getANTsRData( 'r30' )
image_similarity(x,y, type="MeanSquares")
image_similarity(y,x, type="MeanSquares")

image_similarity(x,y, type="Correlation")
image_similarity(y,x, type="Correlation")

# NOT SYMMETRIC
image_similarity(x,y, type="MattesMutualInformation")
image_similarity(y,x, type="MattesMutualInformation")

# NOT SYMMETRIC
mutual_information(x, y)
mutual_information(y, x)

image_similarity(x,y, type="ANTSNeighborhoodCorrelation")
image_similarity(y,x, type="ANTSNeighborhoodCorrelation")
image_similarity(x,y, type="JointHistogramMutualInformation")
image_similarity(y,x, type="JointHistogramMutualInformation")
image_similarity(x,y, type="Demons")
image_similarity(y,x, type="Demons")

muschellij2/extrantsr documentation built on June 3, 2024, 7:04 p.m.