Realistic anatomical model reconstruction and high-fidelity flow computation in stented coronary arteries for the analysis of in-stent thrombosis

Figures

• 

  Figure 1. Reconstruction of a silicone model with a thrombus proximal to the stenosis site, using the conventional and the REFINED approaches (A,B). The location of the thrombus volume is indicated by red shading in panels C and D, while panels E, F, G and H show the streamlines and SR distribution in the two reconstruction models. It is apparent that, at the thrombus site, there are malapposed struts that lead to high and heterogeneous SR regions, which are depicted only by the REFINED method. Panel I shows the SR distribution at the thrombus and non-thrombus sites across the entire set, Panel J shows the percentage of the volume exposed to different levels of SR (<300 s−1, >1000 s−1 and >2000 s−1) in the two cases where the thrombus was present at the location of the stenosis, and in the four cases where the thrombus was present proximally to the stenosis, across the two reconstruction approaches.

• 

  Figure 2. 3D model of a case with a bifurcation lesion treated using the TAP technique, reconstructed with the conventional (A) and REFINED (B) methods. The insert panels show the region of high shear rate (>1000 s−1) at the time of peak flow.

• 

  Figure 3. Methodology implemented to measure the ESS in the models reconstructed by the conventional and the REFINED approaches. Irrespective of the strut apposition status in the conventional approach, the 3D footprint of the struts was used to identify the closest point of the lumen-stent model, and this was used to define two squares: the first with an edge equal to the strut thickness and the second with an edge of 400 μm (A,C,E). The same approach was also followed in the REFINED approach for the malapposed and embedded struts (B,D), while in the case of apposed struts, the two squares were drawn around the stent point that was closer to the 3D footprint of the struts (F) The ESS and multidirectional ESS were then measured in the small square and in the area defined between the large and small squares. In both approaches, the SR was measured in a sphere with a diameter of 320 μm centred on the 3D location of the strut, irrespective of the strut apposition status (G,H,I,J).

• 

  Figure 4. Conventional and REFINED methodologies for coronary artery reconstruction. In the conventional approach, the lumen borders in the non-stented segment and the stent borders in the stented segment are annotated in OCT images (A,C), and then placed perpendicular to the lumen centreline, which is extracted from two angiographic projections (B), to generate a point cloud (C′) that provides the basis for the generation of the 3D lumen-stent geometry surface (C′′). The REFINED approach involves the annotation of the lumen borders in the native and stented segments (D) and the identification of the struts in the stented segment (E) The lumen and stent borders are placed on the lumen centreline extracted from two angiographic projections and the former are used to generate the basis for the lumen silhouette (D′) first, and ultimately the lumen surface (D′′) while, the latter leads to the stent skeleton (E′), which provides the basis for the construction of the stent geometry (E′′). The lumen and the stent models are then combined to generate the final lumen-stent reconstruction (F).