Our powerful registration algorithms allow to fuse pre-interventional imaging such as multi-parametric MRI and (PSMA-)PET with freehand trans-rectal ultrasound (TRUS) for improved prostate cancer screening or targeted biopsies. To deal with these challenging imaging datasets, which are subject to various complex deformations, a number of image analysis technologies are paired with sophisticated fused visualization and interaction.
The ImFusion Suite offers a variety of tools allow for calibration, 3D reconstruction, and registration of freehand ultrasound acquisitions. Both intensity-based and feature-based registration are readily supported by the software. For a powerful intensity-based registration, a statistical deformation model [1] can be used to initialize the non-linear transformation according to commonly pronounced deformation modes for subsequent intensity-based image alignment. The definition of anatomical landmarks together with tools for blending, superimposition and side-by-side visualization enable fusion of e.g. trans-rectal ultrasound acquisitions with an MRI scan. A combination of available anatomical clues and powerful image similarity metrics lead to a robust final alignment of the data despite its difficult appearance and geometrical inconsistency.
[1] O. Zettinig, J. Rackerseder, B. Lentes, T. Maurer, K. Westenfelder, M. Eiber, B. Frisch, N. Navab.
Preconditioned Intensity-Based Prostate Registration using Statistical Deformation Models
Axial (top) and coronal (bottom) slices of prostate MRI (greyscale) and TRUS (red overlay) after deformable image registration. Yellow points indicate anatomical landmark locations, and the white line shows a manual segmentation of the gland in MRI.