Materials and Methods

A total of 30 consecutive patients who were scheduled for any gynecologic surgery and were willing to enroll in the study were prospectively recruited. Only those ovaries assumed to be normal after both USG and the entire operation were analyzed (n = 46). Of the 30 women, with the exception of 1 who underwent vaginal hysterectomy and unilateral salpingo-oophorectomy, the remaining patients were treated with total abdominal hysterectomy and unilateral or bilateral salpingo-oophorectomy. The reasons for hysterectomy were myoma (n = 14), ovarian pathology (n = 6), both myoma and ovarian pathology (n = 2), endometrium cancer without ovarian involvement (n = 2), endometrial hyperplasia (n = 2), persistent premenopausal bleeding (n = 2) or postmenopausal bleeding (n = 1), and stress incontinence (n = 1). Any ovaries bearing a follicle ≥10 mm in diameter or cyst noticed either under USG (n = 8 ovaries, having any degree of echogenicity) or during the surgery (n = 6 ovaries) were excluded.

All 2D and 3D examinations with endocavitary probe (5-9 MHz) were performed by a single physician (Z.Y.) at a maximum of 24 hours prior to the operation with Voluson e (GE Healthcare, Istanbul, Turkey). The 2D OV was estimated with 3 available dimensions, namely maximal longitudinal (a), anteroposterior (b), and transverse (c) diameters, as previously reported. The 3D OV was processed at the same visit with virtual organ computer-aided analysis imaging program using plane A and 60-degree rotational steps.

Following oophorectomy with laparotomy, the true OV was calculated immediately in the operating room by different physicians (G.B. or M.C.S.) who were blinded to estimations of the observer performing the USG assessment. Initially, a sterile sperm tube was filled with 0.9% sodium chloride (NaCl) up to a highlighted marker, which was nearly at midtube. After inserting the ovary into the tube, the amount of liquid passing over the marker was aspirated with an injector with 10 mL capacity. The true OV was accepted as the measured volume inside the injector that was required to drop the liquid level back to the highlighted marker on the tube.

The intraobserver and interobserver reliability was analyzed with another cluster of 24 ovaries. The first observer acquired 2 volumes for each ovary with Archimedes’ principles. Another observer performed a second analysis of the volumes acquired by the first observer. Each observer was blinded to the calculation of the other.

The normality of the distribution of the variables was analyzed by Kolmogorov–Smirnov test. Pearson test was used for correlation analyses. The term “mean bias” was used to define the average difference between true OV and the estimated volume by either 2D or 3D USG. Upper and lower limits of agreement describe “mean bias + 1.96 × SD” and “mean bias – 1.96 × SD,” respectively. Bland-Altman plot of differences was figured, as reported earlier. Intraobserver and interobserver reliability was assessed by calculating the intraclass correlation coefficients for each index. Data analysis was performed using the Statistical Package for Social Sciences, version 13.0 (SPSS Inc, Chicago, IL). Values of P < .05 were considered statistically significant; π was defined as 3.14.

An approval from the institutional review board is available, and written consent was received from all patients.