An objective volumetric method for assessment of ovarian follicular and luteal vascular flow using colour Doppler ultrasonography

Our goal was to develop an objective computer-assisted volumetric method of assessing vascular flow from colour Doppler ultrasound data of ovarian structures recorded by free-hand movement. We hypothesized that a vascularity index (ratio of the region of blood flood to the region of ovarian structure) obtained from the three-dimensional volumetric analysis would be more precise (less variable) than conventional two-dimensional analysis of single images in estimating the functional status of the preovulatory follicles and corpus luteum. Doppler ultrasound cineloops of water buffaloes (Bubalus bubalis; n = 22) ovaries were recorded daily from 12 h before GnRH treatment to four days after ovulation. Cineloops were processed using Fiji and Imaris software packages for segmenting the area (two-dimensional analysis) and the volume (three-dimensional analysis) occupied by the blood-flow and associated tissue to calculate the vascularity index. For volumetric measurement, all images in a cineloop were used (i.e., no a-priori selection of images) while for two-dimensional analysis, three images from the region with apparent maximum vascularity were selected. The volumetric method was verified with theoretical ellipsoidal volume of the follicle (r = 0.96 P < 0.01) or corpus luteum (r = 0.58 P = 0.02). The variability in the follicular vascularity index among animals was lower using the volumetric method than two-dimensional analysis (0.018 ± 0.002 vs 0.030 ± 0.005, P < 0.01), while the variability for CL vascularity was similar between methods (P = 0.23). An increase in the follicular vascularity index was detected at 12 h after GnRH treatment using both methods (two-dimensional: 0.030 ± 0.008, P < 0.01; three-dimensional: 0.016 ± 0.006, P < 0.02). Buffaloes that ovulated tended to have a greater increase in 3D vascularity index than non-responding buffaloes (P = 0.06); the two-dimensional method was not able to detect these changes. Using the three-dimensional method, a moderate positive correlation (r = 0.59; P = 0.02) was evident between the follicular vascularity index at 14–16 h after GnRH treatment and follicular diameter. In conclusion, an objective volumetric method for assessing relative ovarian blood flow changes was developed using Doppler ultrasound cineloops recorded by free-hand movement. The 3-dimensional method eliminates the need for a-priori selection of images and is more precise as a result of decreased technical variability.