Virtual-versus-Real Implantation of Flow Diverters: Clinical Potential and Influence of Vascular Geometry

Left columns, Visual comparison between the 3DRA geometries post-FDS implantations (yellow surface) and pre-FDS implantations (gray surface) for 8 patients. The post- and pre-FDS implantation vessel boundaries were manually registered by using ParaView software. Middle columns, Radius (r) of the stents versus the position (s) along the centerline of the stents (the origin is taken at the distal end). The blue (red) line represents the radius of the virtual stent based on 3DRA pre- and (post-)stent acquisitions, while the gray area shows the minor-major radius range (measured on conebeam CT) of the real stent cross-section based on conebeam CT poststent acquisition. Right columns, Visual comparison between real and virtual FDS implantations. The black lines are the wires of the virtual stents deployed in the 3DRA poststent geometry. The semitransparent purple surface represents the radiopaque wires of the real stents segmented from conebeam CT. The implantations labeled in red required additional manipulation to improve the stent apposition.

Virtual-versus-measured mismatch of the FDS radius [Δr / r = 2 (r_vir − r_mes) / (r_vir + r_mes)] and length [ΔL/L = 2 (L_vir − L_mes) / (L_vir + L_mes)]. A, ΔL/L versus Δr/r for the virtual deployment based on the geometric (patient-specific) model. The virtual deployments represented by the blue (red) squares are based on the vessel radius computed from segmented 3DRA pre- (post-)FDS implantation. The former were corrected by 6% (cyan boxes) to take into account the average radial difference with the measured stent radius in the virtual deployments. The straight black line represents the linear trend (R² = 0.67) with 0.027 ± 0.016 (ΔL/L intercept) and −1.43 ± 0.20 (slope). B, ΔL/L versus Δr/r for a cylindric deployment based on the average radius of the 3DRA pre-FDS (green boxes). The straight black line represents the linear trend (R² = 0.71) with 0.20 ± 0.02 (ΔL/L intercept) and −1.56 ± 0.23 (slope). The filled (empty) boxes correspond to the implantations without (with) additional delivery manipulation (eg, case 17 in Fig 4). C1–2 and D1–2, The virtual-versus-measured FDS radius and length mismatches, respectively. The analysis in C1 and D1 takes into account all the cases while the analysis in C2 and D2 is restricted to the 12 implantations without additional delivery manipulation. The color code is the same as that in the subsets A and B.

Additional manipulation intended to improve the stent apposition in case 17. Image (A) represents the DSA preimplantation. On 3 consecutive roadmap captures (B1–3), the proximal end of the stent was pushed forward by using the tip of the microcatheter to improve the expansion of the stent within the landing zone. This manipulation induced a visible shortening of the device highlighted by the red dot (initial position of the proximal end of the stent just after release) and the white arrow in B3. C1 and C2, Unsubtracted captures just after implantation, respectively, with and without contrast agent injection.