| Popis: |
Until now assessment of regional wall motion abnormalities has relied on either a visual wall motion scoring (WMS) system or tedious off-line border analysis. We explored the potential of a new AQ approach for defining regional WMA. This approach processes RF and integrated backscatter data on-line and displays the endocardial motion as it moves from diastole through systole in a color-encoded manner. Triggered by R wave, as the LV contracts, the automatically extracted border is given designated colors for each frame. As the wall moves during the first few millisec (first frame) it is given one color; then as it moves further during the subsequent frames, various color layers are added on-line. If the wall moves outwards (dyskinesis), a bluish-gray color is displayed. The thickness of a given color layer represents the magnitude of excursion during that frame. The overall thickness of all layers represent the overall wall motion from diastole to systole. By looking at the end-systolic frame alone, the extent of regional wall motion and the temporal characteristics can be ascertained. We employed this color-encoded AQ approach in 17 patients with CAD. WMS, based on conventional methods was derived by an experienced echocardiographer from a total of 172 segments from 2D Echo images alone. This was compared to WMS derived by another observer from the end-systolic frame of color-encoded AQ. In addition, inter-observer variability for each method was evaluated. Results Mean WMS was 1.4 ± 0.8 by 2DE alone, and 1.4 ± 0.7 by color-encoded AQ. Correlation between Color-encoded AQ (y) and 2DE WMS (x) was y = 0.6x + 0.6, r = 0.71, P l 0.001. Inter-observer concordances were: 83% for 2DE: r = 80; 88% for Color-encoded AQ: r = 0.86. Besides displaying the overall extent of motion, color-encoded AQ allowed easy appraisal of the temporal abnormalities in regional contraction. We conclude that color-encoded AQ offers for the first time an on-line method for semi-quantitative evaluation of spatial and temporal abnormalities in regional LV function with less interobserver variability. |
