The Heart Valve Society

Back to 2017 Program


3D Printed Modeling Of Mitral Valve Apparatus From Multi-modality Imaging
Marija Vukicevic, Eleonora Avenatti, Kinan Carlos El-Tallawi, Su Min Chang, Stephen H. Little.
The Methodist DeBakey Heart & Vascular Center, Houston, TX, USA.

OBJECTIVE: We have previously reported on the feasibility of 3D printing the mitral valve (MV) leaflets using material blends that approximate mechanical properties of leaflet tissue. We sought to reconstruct the entire MV apparatus and the left ventricle (LV) from CT and 3D transesophageal echocardiographic (TEE) imaging datasets and to use them as tools for patient specific valve interventions.
METHODS: CT (Force, Siemens) and 3D TEE (IE33-Philips) DICOM diastolic and systolic images were exported into segmentation software (Mimics, Materialize). A tagging algorithm was developed to enhance the segmentation of the MV leaflets and chordae complex using color-coding of each chordae branch and papillary muscles. Digital co-registration of those geometries from different cardiac cycle phases was performed and a patient-specific MV was replicated using multi-material 3D print blends which incorporated differentiated regional material stiffness. A bench-top MitraClip (Abbott) procedure was performed within this patient-specific MV apparatus model.
RESULTS: Patient-specific MV apparatus was reconstructed using 3D TEE or CT datasets. However, myocardial reconstruction was possible only from CT data. The feasibility of 3D printed modeling of the entire MV apparatus and LV from 3D TEE images was demonstrated. Echo-based chordal segmentation was facilitated by the novel tagging algorithm - the co-registered systolic chordae enhanced the diastolic segmentation and the anatomical justification. The CT-based left heart was modeled along with the echo-based MV geometry of the same patient. The MV modeled from 3D TEE, co-registered with the CT-based LV model demonstrated the feasibility of the multi-modality imaging data integration into a single patient-specific model of the entire MV apparatus. This multi-material patient-specific model was used to perform a benchtop MitraClip implantation which demonstrated secure device fixation and targeted leaflet apposition (figure).
CONCLUSIONS:The geometry of a patient-specific MV apparatus can be accurately replicated using 3D printed methods. Digital co-registration of anatomical elements from different imaging phases and modalities could further improve their accuracy and could serve as a clinical tool for the simulation and planning of heart interventions.


Back to 2017 Program