The computerized linear measurement cross-hairs were used to ensure optimum precision. We found that observing the right ventricular wall motion on two-dimensional image allowed for better definition of the right ventricular wall than static images would allow. The M-mode technique, as described by Lesser et al,s was performed to corroborate these measurements. All measurements were obtained at end-diastole (onset of the QRS complex by simultaneous electrocardiogram) at approximately the level of the mitral valve similar to the T, and Tf positions described by McKenna et al.” The RVH measurements were obtained on at least five separate cycles.
The average was computed and used as the RVH score. We used two independent scorers, each of whom was a board-certified cardiologist, who were blinded to the patients history, AI and to the other scorers RVH score. Statistical analysis was performed to evaluate reproducibility. We used 5 mm as the cutoff for normality of the right ventricle, as has been previously noted in the literature. To be defined as RVH, the subjects had to have an RVH score greater than or equal to 6 mm by both scorers. In addition, electrocardiograms of each patient were read by the two reviewers for evidence of RVH.
The data were analyzed utilizing the Students t test. The Kolmogorov-Smirnov test of normality and the test for homogenous variance were performed to ensure validation of the Student’s t test.