Related Experiment Video
Updated: Mar 24, 2026
Evaluation of Hepatic Glucose Production in a Polycystic Ovary Syndrome Mouse Model
Published on: March 5, 2022
Diagnosis of Polycystic Ovary Syndrome With Predictive Modeling of Select Clinical Features
Adam T Evans1, Eeshaan Rehani1, Bailey Smith1
1Department of Obstetrics and Gynecology, University of Rochester Medical Center, Rochester, New York; Division of Reproductive Endocrinology and Infertility, University Hospitals/Case Western Reserve University School of Medicine, Beachwood, Ohio; Department of Physiology and Biophysics, Englander Institute for Precision Medicine, Institute for Computational Biomedicine, and The Ronald O. Perelman and Claudia Cohen Center for Reproductive Medicine, Weill Cornell Medicine, New York, New York; and Meinig School of Biomedical Engineering and Division of Nutritional Sciences, Cornell University, Ithaca, New York.
Objective:
To determine whether a limited set of ultrasonographic, biochemical, and clinical features are sufficient to accurately predict polycystic ovary syndrome (PCOS) diagnosis.
Methods:
Transvaginal ultrasound images and available clinical data for participants with PCOS (n=101) and controls (n=50) were used for this multicenter, retrospective pilot study. Diagnosis of PCOS was defined by the 2023 International Evidence-Based Guideline. Controls had no diagnostic features of PCOS. Differences in demographic (age, body mass index [BMI]), ultrasonographic (ovarian volume, follicle number per ovary, follicle number per single cross section), biochemical (sex hormone binding globulin, total testosterone, free androgen index, bioavailable testosterone), and clinical (follicle-stimulating hormone, luteinizing hormone, estradiol, anti-müllerian hormone (AMH), age at menarche, minimum self-reported menstrual cycle length in the past year, maximum self-reported menstrual cycle length in the past year, Ferriman-Gallwey hirsutism score) features between groups were assessed with Mann-Whitney U tests. A logistic regression model was trained to predict PCOS diagnosis using subsets of ultrasonographic, biochemical, and clinical variables. Model performance was evaluated with area under the receiver operating characteristic curve (AUROC) and F1 score measures.
Results:
Anti-müllerian hormone alone predicted PCOS diagnosis with relatively good diagnostic accuracy (AUROC 0.884, F1 score 0.807). Integration of AMH and ovarian volume improved model performance (AUROC 0.906, F1 score 0.811). Integration of all ultrasonographic, biochemical, and clinical features yielded a high-performing model with excellent diagnostic accuracy for PCOS (AUROC 0.991, F1 score 0.811). Refinement of the model to a limited set of readily obtained variables, including AMH, ovarian volume, hirsutism score, and maximum self-reported menstrual cycle length, yielded a model with strong performance (AUROC 0.982, F1 score 0.805).
Conclusion:
A minimum combination of ovarian volume with AMH and examination/history-derived end points can accurately predict PCOS diagnosis with performance comparable to that of a combination of all ultrasonographic, biochemical, and clinical features. This may streamline diagnostic workflows, thereby reducing clinical burden.
Parent Protocol Clinical Trial Registration:
ClinicalTrials.gov, NCT01859663, NCT01927471, NCT03306849, NCT01927432, NCT03547453.
