SEMIAUTOMATIC SEGMENTATION OF GATED BLOOD-POOL EMISSION TOMOGRAPHIC-IMAGES BY WATERSHEDS - APPLICATION TO THE DETERMINATION OF RIGHT AND LEFT EJECTION FRACTIONS

Tomographic multi-gated blood pool scintigraphy (TMUGA) is a widely av ailable method which permits simultaneous assessment of right and left ventricular ejection fractions. However, the widespread clinical use of this technique is impeded by the lack of segmentation methods dedic ated to an automatic analysis of ventricular activities. In this study we evaluated how a watershed algorithm succeeds in providing semi-aut omatic segmentation of ventricular activities in order to measure righ t and left ejection fractions by TMUGA. The left ejection fractions of 30 patients were evaluated both with TMUGA and with planar multi-gate d blood pool scintigraphy (PMUGA). Likewise, the right ejection fracti ons of 25 patients were evaluated with first-pass scintigraphy (FP) an d with TMUGA. The watershed algorithm was applied to the reconstructed slices in order to group together the voxels whose activity came from one specific cardiac cavity. First, the results of the watershed algo rithm were compared with manual drawing around left and right ventricl es. Left ejection fractions evaluated by TMUGA with the watershed proc edure were not significantly different (p=0.30) from manual outlines w hereas a small but significant difference was found for right ejection fractions (p=0.004). Then right and left ejection fractions evaluated by TMUGA (with the semi-automatic segmentation procedure) were compar ed with the results obtained by FP or PMUGA. Left ventricular ejection fractions evaluated by TMUGA showed an excellent correlation with tho se evaluated by PMUGA (r=0.93; SEE=5.93%; slope=0.99; intercept = 4.17 %). The measurements of these ejection fractions were significantly hi gher with TMUGA than with PMUGA (P<0.01). The interoperator variabilit y for the measurement of left ejection fractions by TMUGA was 4.6%. Ri ght ventricular ejection fractions evaluated by TMUGA showed a good co rrelation with those evaluated by FP (r = 0.81; SEE = 6.68%; slope = 1 .00; intercept = 0.85%) and were not significantly different (P = 0.42 ). The interoperator variability with TMUGA was 6.7% for the right ven tricle. Thus, the watershed algorithm proposed is an efficient segment ation tool for the semi-automatic analysis of right and left ventricul ar ejection fractions by TMUGA. Further studies are necessary to check whether this procedure can be used to evaluate ventricular volumes an d cardiac outflow.