Variation in Form and Function of the Non-Pregnant Uterus

This study aimed to develop a framework to quantify variation in uterine morphology using magnetic resonance imaging (MRI)-based segmentation and Principal Component Analysis (PCA), and to assess whether these anatomical differences influence electrical signal propagation through patient-specific simulations. The framework was applied to a novel pilot dataset of participants with (n = 2) and without (n = 5) medically confirmed endometriosis, to explore potential group-level differences in both morphology and electrophysiology. Uterine position and overall shape were assessed using 2D MRI measurements, including angle of version, flexion, and internal shape. Positional variations such as anteversion, retroversion, and retroflexion, along with arcuate internal shapes, were observed across the cohort; however, no consistent group-level patterns were observed. These assessments were complemented by 3D segmentation, enabling quantitative analysis of uterine volume. Participants with endometriosis exhibited significantly larger total uterine volumes (mean: 127.2 cm³, range: 120.9–133.5 cm³) compared to those without (mean: 74.9 cm³, range: 60.3–84.2 cm³, p = 0.002). Significant differences were also observed in myometrial (97.8 vs. 58.4 cm³, p = 0.018) and endometrium/uterine cavity volumes (9.3 vs. 4.6 cm³, p = 0.031). This study demonstrates the value of combining 3D quantification with traditional 2D assessments, highlighting discrepancies that may arise from reliance on 2D methods alone. PCA was used to assess variation in uterine shape by reducing the complex geometry of segmented meshes. Over 95% of total shape variance was captured within the first three modes. Mode 1 reflected elongation and orientation along the fundus-to-cervix axis, while Modes 2 and 3 captured curvature, bulging, and torsional deformation. To the best of our knowledge, this is the first study to apply such quantitative 3D shape analysis to the non-pregnant uterus, providing a scalable alternative to traditional 2D morphological assessments. Electrical simulations using a smooth muscle cell mathematical model, adjusted for human-relevant hormonal levels, revealed correlations between specific anatomical features and simulated signal propagation. However, no statistically significant group-level differences were observed, suggesting that morphology alone may not fully explain functional alterations reported in endometriosis. Together, this framework provides a foundation to investigate variation in the form and function of the non-pregnant uterus.