Phantom Preparation
A silicon coronary artery phantom simulating main branch (MB) and side branch (SB) was prepared using a 3-D printer (Projet 3510 SD).
An angle between MB and SB was 45 degree and wall thickness of the phantom was 1 mm.
The MB diameter was 4.5 mm tapering to 3.5 mm and the SB diameter was 3.0 mm.
Bifurcation stenting was performed by an experienced interventional cardiologist (Y.H.K,
8 years of experience) using a drug eluting stent (Sience Prime,
Abbott).
Crush,
Cullotte,
and T-stenting technique with final kissing balloon dilatation were performed in 4,
3 and 2 phantoms,
respectively.
CT Imaging - Phantom
Total of 9 deployed phantoms were imaged using both micro-CT (SkyScan 1172,
Bruker,
Beligium,
acquisition time 40 min,
FOV 50 mm,
spatial resolution <1µm) and second generation dual-source CT scanner (Somatom definition; Siemens Medical Solution,
Forchheim,
Germany).
The micro-CT images were reconstructed with 27-um slice image and exported to the off line software (CTAn,
Meshlab).
Clinical CT images were reconstructed with OO-mm slice image and exported to the off line software (Syngovia).
analyzed using CTAn (Bruker,
Beligium) and MeshLab (freeware),
and conventional CT images were analyzed using Syngovia (Siemens Medical Solutions,
Erlangen,
Germany).
CT Analysis
Parameters including crushed segment,
floating strut,
the location of the crushed segment,
unintended gap and overlapped segment were defined in our study (figure 1).
The definition of crushed segment was overlapping of the two layers of crushed SB stent struts in addition to a layer of MB struts on micro-CT (figure 2).
Floating strut was defined as a malapposed and protruded strut into the lumen causing luminal narrowing (figure 2).
The location of crushed segment was divided to two types; the angle between the center of MB and crushed segment from -45 degree to 45 degree defined as central and otherwise eccentric.
When the center of crushed segment lined up straight with the SB stent it was classified as central and other than central,
it was classified as eccentric (Figure 3).
Unintended gap was defined as absence of stent strut between the MB and SB stent (Figure 4).
In phantoms of Culotte stenting,
additional parameter including overlapped segment was applied and its definition was circumferential overlapping of one layer of MB strut and one layer of SB strut.
Several parameters were measured in both micro-CT and dual source CT (figure 1).
The length and perimeter of crushed segment,
the minimal luminal area at the level of crushed segment,
and SB ostium were measured.
In phantoms with floating struts,
the length of floating strut and the minimal luminal area in MB and SB was measured.
Additionally,
the area of SB at non-crushed segment 1-cm distal to SB ostium was measured.
In phantoms of Culotte stenting,
the longest length and perimeter of overlapped segment,
and the minimal luminal area at the level of overlapped segment were measured.
The location of crushed segment,
the presence of floating strut and unintended gap were compared between micro-CT and dual source CT.
All CT image analyses were performed in a blinded manner by two independent observers.
Statistical analysis
Continuous variables were expressed as mean ± standard deviation,
and categorical data were presented as percentages.
To evaluate the agreement of measured parameters between micro-CT and dual source CT,
Bland-altman plot was used.
A statistical software package (MedCalc Software,
Mariakerke,
Belgium) was used to analyze the data.