RESUMEN
M-mode echocardiographic measurement of the left ventricular mass is inaccurate when the hypertrophy is asymmetric and the ventricule very deformed. The routine calculation of the mass is based on several hypotheses, verified in normal and hypertensive subjects: the standard error (SEE) is 30-40 g, 10-15% (r > 0.9). The standard deviation of inter-examination differences (> 25 +/- 30 g with a variation coefficient c = 10-15%) makes it difficult to appreciate variations of mass in a given patient. Two-dimensional echocardiographic measurement of left ventricular mass requires the use of geometric formulae which have not been validated in the cardiomyopathies. The absence of a consensus on the models used has favorized the use of MRI and of ultrafast computed tomography. MRI measurement of mass has been validated in normal and ischaemic hearts (r > 0.97, SEE < 8 g, c = 15%) but the times of acquisition are long. Using ultrafast CT, not universally available, this measurement has been validated in vivo and in vitro, including in cardiomyopathy for which the SEE is low (6%) and reproducibility excellent (c = 4-8%), comparable with results in normal subjects.