3.2 Annular measurements and reproducibility
A comparison of the aortic annular measurements between MSCT and both 3D TEE methods are shown in Table 2 . Both 3D TEE methods highly correlated with MSCT for annular diameters (semi-automated r =0.96, p<0.0001 and manual r=0.84, p<0.0001) Figure 3 and for annular areas (semi-automated r=0.96, p<0.0001 and manual r=0.82, p<0.0001) Figure 4 . MSCT and semi-automated 3D TEE annular area measurements predicted similar TAVR prosthesis size except in only 1 patient where the semi-automated 3D TEE annular area size actually suggested a bigger TAVR prosthesis size, compared to MSCT annular area size. On the other hand, MSCT and manual 3D were discordant in 5 patients in predicting TAVR annular size. Of these, manual 3D TEE predicted a smaller TAVR prosthesis in 3 patients, compared to MSCT annular area measurements. Inter-observer reproducibility was high for both 3D TEE methods for annular diameters (semi-automated r=0.96, p<0.001 and manual r=0.91, p<0.001) Figure 5 and annular areas (semi-automated r=0.96, p<0.0001) and manual was 0.95 p<0.0001)