A new multimodal imaging technique for virtual planning of mitral valve repair

Introduction: The geometric shape of the mitral valve annulus has an important impact on the results of the mitral valve repair. Recent research has shown a broad range of individual variability in anatomy of mitral valve apparatus both in normal and in diseased states, i.e. saddle shape and (a-)symmetrical annulus dilatation. Currently, no accurate method exists for preoperative assessment of mitral annulus shape. We evaluated in this context the feasibility of a hybrid imaging technique for individualized visualization of mitral valve annulus geometry prior to operation.

Methods: A total of 15 patients underwent transthoracic 3DE (Sonos, 7500) and cardiac magnetic resonance imaging (MRI; Siemens, 1.5 T).

Results: The 3DE and MRI data sets were post-processed using a newly developed software (MITK, DKFZ, Heidelberg), which was implemented with an algorithm for segmentation of the mitral valve annulus. Additionally, the MRI data sets were used for stereolithographic reconstruction of every individual heart including the exact geometrical reconstruction of the mitral valve apparatus. The stereolithographic models were digitally scanned and visualized as virtual 3D images. The segmentated annulus of each patient was then virtually implanted into the 3D image stimulating thereby a virtual mitral valve annuloplasty.

Conclusions: Our data show that, using multimodal imaging techniques, an individual-based planning of the complex cardiac procedures, i.e. mitral valve annuloplasty can be performed on a virtual model of the patient's heart. We are confident that our data has great potential for clinical use and can improve the outcome of operations.