Research Article

Evaluation of a Nonrigid Motion Compensation Technique Based on Spatiotemporal Features for Small Lesion Detection in Breast MRI

Table 3

Areas under the ROC curves for morphological features (geometric moments 𝐼 1 to 𝐼 6 ) using FLDA. The rows represent the motion compensation as given in Table 1. Numbers in boldface show the best results.

Feature typeArea under the ROC curve (%)
010203040506070809

𝐼 1 64.5 54.464.760.363.267.6 65.864.166.2
𝐼 2 56.358.056.754.269.560.763.766.859.7
𝐼 3 56.378.8 65.558.059.067.268.161.662.2
𝐼 4 56.355.961.863.456.358.263.959.254.0
𝐼 5 56.352.958.866.264.564.564.561.859.2
𝐼 6 56.356.363.759.960.167.266.064.366.4