Echocardiographic evaluation of responses to left ventricular volume loading by principal components and nomographic...

Changes in performance of left ventricles (LV) with volume overloads are difficult to determine by conventional clinical methods. This information, however, is important for the proper timing of therapeutic interventions to preserve LV function. Seven size-adjusted (by regression with end-diastolic diameter, EDD) LV performance parameters from 100 normal echocardiograms (echos) were entered into a principal component analysis (PCA). Two factors (linear combination of the 7 parameters) were obtained from the analysis. Prediction limits (95%) about these two factors used in combination, correctly classified 92% of the normal echos. More detailed analysis of the two PCA factors revealed that the majority of the variability within the factors was explained by size-adjusted parameters resultant from the EDD posterior wall (factor 1) and EDD septal excursion (factor 2) regressions, respectively. Plots of the 95% prediction limits about these two regression lines provided nomograms. These nomograms used in combination, correctly classified 95% of the normal echos. When the performance parameters of 64 volume loaded ventricles were evaluated by PCA, four groups were identified. Ten ventricles (16%) were hypokinetic, 29 (45%) were hyperkinetic, 23 (36%) were nomokinetic, and 2 (3%) could not be classified. These classifications were supported by significant between group differences of shortening fraction, velocity of circumferential shortening, and velocity of circumferential expansion. Nomographic classification of the same volume loaded hearts was in excellent (94%) agreement with PCA classification. Nomographic analysis (derived from the PCA) is offered as a less complex, more clinically applicable echo method for evaluating LV performance of volume loaded hearts.