The visualization of bioprosthesis leaflet morphology might help to better understand the underlying mechanism of dysfunction in degenerated aortic bioprosthesis. Because today such visualization of bioprosthesis leaflet morphology is intricate to impossible with other imaging techniques, we hypothesized that the processing of multi-detector CT images would allow better visualization of the prosthetic valve leaflets after biological aortic valve replacement. The purpose of our study was to prospectively evaluate patients with a degenerated aortic bioprosthesis, waiting for reoperation, by using 64-slice CT to evaluate prosthetic leaflets morphology. A semi-automatic segmentation of pre-operative tomodensitometric images was conducted, using 2 different implementations of the region growing algorithm. Here we report all segmentation steps (selection of the region of interest, filtering, segmentation). Studied degenerated aortic bioprostheses were represented by two Carpentier-Edwards Supra Annular Valve (porcine leaflets), one Edwards Perimount (pericardial leaflets) and one Medtronic Mosaic (porcine leaflets). Both segmentation methods (Isotropic Region Growing and Stick Region Growing) allowed a semi-automatic segmentation with 3D reconstruction of all bioprosthetic components (stent, leaflets, degeneration/calcifications). Explanted bioprosthesis CT images were also processed and used as reference. Segmentation results were compared by means of quantitative criteria. Semi-automatic segmentation using region growing algorithm seems to provide an interesting approach for the morphological characterization of degenerated aortic bioprostheses. We believe that in the next future CT scan images segmentation may play an important role to better understand the mechanism of dysfunction in failing aortic bioprostheses. Moreover, bioprostheses 3D reconstructions could be integrated into preoperative planning tools to optimize valve-in-valve procedure.
Read full abstract