Skip to main content

Table 1 Dice (upper triangle) similarity scores and correlation coefficients (lower triangle) between atlases build from different initializing samples

From: Performance of single and multi-atlas based automated landmarking methods compared to expert annotations in volumetric microCT datasets of mouse mandibles

 

Sample 1

Sample 2

Sample 3

Sample 4

Sample 1

 

0.997

0.997

0.997

Sample 2

0.999

 

0.998

0.997

Sample 3

0.999

0.999

 

0.997

Sample 4

0.999

0.999

0.999

 
  1. DICE similarity is calculated as the ratio of twice the intersection of two images divided by sum the two images, with score of 1 representing two identical images. Four samples randomly chosen from the study population to initiate the atlas building process