Sigel H, Nechwatal W, Stauch M
Z Kardiol. 1981 Oct;70(10):742-7.
In 17 patients without evidence of regional wall motion abnormality a determination of the variation of global and regional parameters of two contour delineations was attempted by the same (intraobserver) and by an independent observer (interobserver). In the intraobserver comparison all computed comparison all computed parameters showed excellent correlation coefficients: enddiastolic volume (EDV) r = 0.97, endsystolic volume (ESV) r = 0.94, ejection fraction (EF) r = 0.93. In the interobserver comparison, the correlations were slightly lower (EDV: r = 0.94, ESV: r = 0.93, EF: r = 0.85). Absolute differences in volumes and ejection fractions were for the EDV 3.0 +/- 11.8 ml (+/- 1 SD), for the ESV 2.6 +/- 4.9 ml, and for EF 1.9 +/- 4.1%. Additionally, the three contour tracings were subjected to an analysis of regional wall motion using nine different methods. With most of the methods applied, the analysis showed areas near valves and the cardiac apex, where small changes in contour tracings resulted in statistically significant differences in wall motion (p less than 0.05). Therefore, using each method described, it was possible to determine, which regions demonstrated the greatest variation in interpretation of quantitative regional wall motion. Finally we conclude, that enddiastolic and endsystolic volumes and ejection fraction can be subjected to repeated estimation by the same or an independent observer with sufficient accuracy.