The accuracy of transrectal real-time ultrasonography (RTU) scanning technique to detect ovarian structures (follicles and corpus luteum) of Serrana goats was compared to the data obtained by observation of ovarian sequential slices. This slicing technique (SLI) was considered as reference method. The laparoscopy and laparotomy techniques were also used for corpora lutea identification. For this purpose the ovaries of 14 females were observed, 7-8 days after ovulation, by transrectal ultrasonography followed by laparoscopic examination. Then ovaries were removed and studied by slicing. In the sliced sections of each ovary (n = 28), follicles and corpus luteum (CL) were identified and counted. CL and follicular diameters were measured using a millimetre scale.The total number of follicles, counted by RTU, was significantly lower than that observed by SLI (P < 0.01). This difference was mainly due to the under estimation of <2 mm follicles category. The correlation coefficient between category data obtained by RTU and SLI methods for the number of follicles ≥3 mm was high (r 2 = 0.95, P < 0.001), which highlights the use of UTR as a potential methodology to study the follicular dynamic of goats.There were no significant differences (P > 0.05) between the average number (mean ± S.D.) of corpus luteum identified per ovary by RTU (0.71 ± 0.75), laparoscopy (0.58 ± 0.71), laparotomy (0.67 ± 0.76) or SLI (0.83 ± 0.76) methods. The accuracy for the identification of ovulation, validated by CL detection on D7-D8 by SLI (100%), was 91.7%, 87.5% and 83.3% by RTU, laparotomy and laparoscopy, respectively. The negative predictive value of RTU, laparotomy and * Corresponding author. E-mail address: jsimoes@utad.pt (J. Simões). Simões et al. / Animal Reproduction Science 85 (2005) [263][264][265][266][267][268][269][270][271][272][273] laparoscopy to verify the absence of a CL in the ovary was 81.8%, 75.0% and 69.2%, respectively. The specificity of all three methods for the CL identification was 100%. No significant differences (P > 0.05) were found in the probability to detect the exact number of CL (0, 1 or 2) counted in each ovary between the RTU (87.5%), laparotomy (83.3%) and laparoscopy (75.0%) methods when compared with the reference method.
0378-The diameter of spherical CL could be estimated with reliability (r 2 = 0.86; P < 0.001). The real-time ultrasonographic scanning proved to be a highly accurate method for detection and measurement of several categories of follicles and CL size in Serrana goats. The results of the present study show that laparoscopy and RTU are similarly reliable techniques for CL detection. However, the RTU represents a non-traumatic technique with advantages to animal welfare both in experimental and reproductive evaluation of the size of ovarian structures.