Click on image to view larger version.

Figure 3 Accuracy of registration: in vivo evaluation. Datasets using the normal animals were evaluated. The image integration was performed based upon either (A) sequential registration of the MEAM LV points with the MRI-based LV surface, or (B) registering the MEAM aorta points with the MRI surface, followed by sequential incorporation of the MEAM LV points with the MRI-based LV surface. As described in the legend of Figure 1, the registration accuracy was defined as the mean distance of either: 1) the complete MEAM LV point dataset to the MRI-based LV surface (red lines), or 2) the complete MEAM aorta point dataset to the MRI-based aorta surface (blue lines). Congruent with the in vitro phantom data, registration based solely upon the LV (A) resulted in an inaccurate local minimal solution which could be rectified by first registering the aorta points (B). Note that even after proper registration (figure in B), many of the MEAM LV points, but not the aorta points, were located beyond the MRI-defined LV endocardial boundary. Abbreviations as in Figure 1.