Quantification of 3-D variations in breast topology during repeat MR setups

A new method has been developed and applied to quantify the geometrical reproducibility of internal structures in the breast in supine position using MRI. Normalized cross correlation is employed to identify corresponding fiducial points in the fat-gland transitions in the breasts. Systematic and random variations in the position of the fiducial points relative to MR visible skin markers are calculated in repeat scans. In this preliminary study, measurements are performed on 7 healthy volunteers. The effect of a breast immobilization cast is studied. The results indicate systematic and random variations in the position of corresponding fiducials of the order of 2 mm. The use of an immobilization cast does not further improve the reproducibility of the topology.