Estimation of breast volume and its variation during the menstrual cycle using MRI and stereology.

Unbiased estimates of breast volume may be obtained in vivo from systematic series of MR images acquired in accordance with the Cavalieri method of modern design-based stereology. The method does not require any assumptions to be made regarding breast shape. If point counting techniques are used to obtain the required breast section areas estimates, 10-15 min analysis (i.e. counting about 250 points on 12 to 16 images) ensures that the contribution of sectioning and point counting to the coefficient of error (CE) on the volume estimate is less than 3%. The methods were applied to measure breast volume in 15 healthy females aged between 22 and 44 years (mean 31.7 years; SD 8.2 years). One subject was studied on every fourth day during two consecutive cycles. The other 14 subjects were studied on three occasions corresponding to menses, ovulation and pre-menses during a single menstrual cycle. Repeat imaging after repositioning on three occasions within a period of 30 min and also at three different times of day for a single subject, both yielded a coefficient of variation (CV) of less than 3% in the estimation of breast volume. ANOVA indicates that there is no significant difference between the mean volume of the left and the right breast (p = 0.294). The mean volume of the left breast is 561 ml (95% confidence interval (CI): 553 ml, 569 ml) and the mean volume of the right breast is 567 ml (95% CI: 559 ml, 576 ml). There are highly significant differences between the three named stages of the menstrual cycle (p < 0.0005), whereby the mean volume at ovulation is 5.5% less than the mean volume at menses (95% CI: 3.0%, 7.9%) and the mean volume pre-menses is 8.1% greater than the mean volume at menses (95% CI: 5.3%, 10.9%). Overall, the volume of each breast varies by an average of 76 ml (95% CI: 61 ml, 92 ml) during the menstrual cycle, which corresponds to 13.6% of the volume at menses (95% CI: 13.3%, 13.8%). No significant interaction was found between the relative volumes of the left and right breast and the stage of the menstrual cycle (p = 0.277), nor between subjects and stages of cycle (p = 0.296). However, a significant interaction was observed between the volume of the left and right breasts in different subjects (p < 0.005). The average difference in the volume of the left and right breasts of all 15 subjects is 39.7 ml (95% CI: 21.3 ml, 58.1 ml), which is 7% of average breast volume and approximately 50% of the average variation in the volume of the breast during the menstrual cycle.