To diagnose and manage breast cancer, X-ray mammography, ultrasound, and magnetic resonance (MR) imaging are used to image the breasts. In this chapter, we present a method for validating finite element (FE) biomechanical models across the entire breast volume. A computational framework was used to generate personalised FE models of the breast to predict the deformations of the breasts under compression. A Fourier-based correlation technique was used to compare localised mismatches between 3D model-warped images and clinical MR images. This technique is illustrated by matching model-simulated compressed images with MR data for a compressed breast phantom and the compressed breasts of two volunteers. The results from this analysis indicate regions of the biomechanical model that do not align well, and thus require improvement. A validated biomechanical image registration tool of this kind will help clinicians to more accurately localise breast cancers, enabling more reliable diagnoses.