Small Lesions Evaluation Based on Unsupervised Cluster Analysis of Signal-Intensity Time Courses in Dynamic Breast MRI

Table 5

Results of the comparison between the two different segmentation methods for the 40 small lesions: the average area under the curve 𝐴_𝑍 and its deviations for 20 different ROC runs using the same parameters but different algorithms’ initializations.

Segmentation method	Area under the curve 𝐴𝑍
Conventional segmentation analysis	0.6207
“Neural-gas” network	0.8388