Ultrasound image attributes of human ovarian dominant follicles during natural and oral contraceptive cycles

Abstract

Background: Computer-assisted analyses were used to examine ultrasound image attributes of human dominant ovarian follicles that developed during natural and oral contraceptive (OC) cycles. We hypothesized that image attributes of natural cycle follicles would quantitatively differ from those in OC cycles and that OC cycle follicles would possess image attributes indicative of atresia.

Methods: Dominant ovarian follicles of 18 clinically normal women were compared using transvaginal ultrasonography for the 7 days before ovulation during a natural cycle (n = 9) or the 7 days before peak estradiol in women using OC (n = 11). Follicles were analyzed using region and line techniques designed to compare the image attributes numerical pixel value (NPV), pixel heterogeneity (PH) and area under the curve (AUC).

Results: NPV was higher in OC cycle follicles with region analysis and tended to be higher with line analysis (p = 0.005 and p = 0.06, respectively). No differences were observed in two other image attributes (AUC and PH), measured with either technique, between natural and OC cycle follicles.

Conclusion: The increased NPV value of OC cycle follicles and lack of differences in PH and AUC values between natural cycle and OC cycle follicles did not support the hypothesis that OC cycle follicles would show ultrasonographically detectable signs of atresia. Image attributes observed in OC cycle follicles were not clearly indicative of atresia nor were they large enough to preclude preovulatory physiologic status in OC cycle follicles.

Figure 1

Growth profiles of natural and OC cycle follicles

Figure 2

Ultrasonographic images of a natural cycle and OC cycle follicles. (a) Natural cycle follicle on Day 0, (b) OC cycle follicle on Day 0.

Figure 3

Graphical representation of NPV, PH and AUC. Mean ± SEM (a, b) numerical pixel value (c,d) pixel heterogeneity and (e, f) area under the curve obtained by region analysis (a, c, e) and line analysis (b, d, f) of natural cycle and OC-cycle follicles.