The reproducibility of endometrial volume acquisition and science measurement with the Vocal-imaging program

Abstract

Objectives: To determine the intraobserver reproducibility of volume acquisition and repeatability of endometrial volume measurements using the VOCAL™-imaging program (Virtual Organ Computer-aided AnaLysis). Methods: Ten three-dimensional (3D) ultrasound datasets of the same endometrium were obtained from one patient during one scan session. For every 3D dataset, the endometrial volume was calculated by four different measurement techniques; three rotational methods and one conventional method. A single observer acquired each dataset and conducted all measurements. For the rotational technique, the manual mode was used to trace the myometrial-endometrial border through three different predefined rotations of 30°, 15° and 9°. Endometrial volume was also measured in a conventional manner by scrolling through one of the multiplanar images and outlining the contour in another. Ten consecutive measurements were taken for each of the four measurement techniques so that 40 endometrial volume readings were obtained for each of the 10 3D-ultrasound datasets. Results: Endometrial volume measurements were more reproducible with the rotational technique. Rotational methods employing a rotation step of less than 30° were associated with coefficients of variation below 1% and intraclass correlation coefficients above 0.94. Within each of the 10 different scans, the homogeneity of the variance of measurements was significantly different according to the method of measurement (P = 0.022 to > 0.001). Post-hoc tests revealed that both the conventional and 30° rotational methods resulted in a significantly smaller mean endometrial volume than both the 15° and the 9° rotational methods. Conclusions: Endometrial volume acquisition is reproducible and endometrial volume measurements are repeatable but they depend upon the technique of volume calculation in so much as rotational methods employing a rotation step of less than 30° are associated with a significantly smaller variance in measurements and a significantly greater mean endometrial volume.