[Left ventricular volume measurement; the accuracy of the measurement].

Volume angiography is a widely diffused technique in hemodynamic laboratories despite several factors which limit the validity of the method. We have evaluated three of these factors: a) indefiniteness in the identification of the contour of the left ventricular cavity on the projected image; b) indefiniteness due to the reproducibility in tracing the contour of the left ventricular cavity by the same operator; c) indefiniteness related to the geometrical assumption (ellipsoid of revolution). The indefiniteness, given as per cent variation from the mean value, has been determined for the diastolic as well as systolic phase. The maximal possible indefiniteness, for diastole and systole, in contours identification in 15 cases has been found to be +/- 11.3% and +/- 24% of the mean volume. Intra observer reproducibility evaluated outlining 30 times the left ventricular contour has been equal to a mean difference from the mean volume of +/- 1.2% for diastole and +/- 4% for systole. The mean difference of the mean volume calculated with the area-lenght method (ellipsoid geometry) from the volume calculated with the Simpson rule has been, on 15 cases, equal to 4% in diastole and 11% in systole. This high level of indefiniteness, mainly for the systolic phase, must be considered before giving a value excessively precise to the volume data obtained with angiography. On the other hand, the method retains an orientative meaning extremely useful in clinical evaluation.