Related Experiment Video
Updated: Jun 13, 2026
Noninvasive Monitoring of Lesion Size in a Heterologous Mouse Model of Endometriosis
Published on: February 26, 2019
Machine learning models for non-invasive endometriosis triage using a laparoscopically and histologically verified
Marina Paula Andres1, Clarisse da Costa Rufino2, Ramanathan Anantharama Krishnan2
1HerHealth.AI, Orlando, FL, USA; Disciplina de Ginecologia, Departamento de Obstetrícia e Ginecologia, Faculdade de Medicina FMUSP, Universidade de Sao Paulo, Sao Paulo, BR, Brazil; Gynecologic Division, Beneficência Portuguesa de São Paulo, Sao Paulo, SP, Brazil.
Machine learning models can help triage women at risk for endometriosis using clinical data. These models, validated in a verified cohort, show promise for non-invasive endometriosis risk assessment.
Area of Science:
- Gynecology
- Medical Informatics
- Machine Learning
Background:
- Endometriosis diagnosis often requires invasive procedures.
- There is a need for non-invasive methods to identify women at risk for endometriosis.
- Structured clinical variables can potentially be used for risk stratification.
Purpose of the Study:
- To develop and validate machine-learning models for non-invasive endometriosis risk triage.
- To utilize structured clinical variables from a laparoscopically and histologically verified cohort.
- To compare clinician-guided and statistically optimized feature selection strategies.
Main Methods:
- Retrospective study of 2546 women undergoing laparoscopic surgery.
- Stratified 10-fold cross-validation with preprocessing, feature selection, MinMax scaling, and SMOTE.
- Performance evaluation using F1-score, recall, PPV, NPV, and AUC-ROC.
Main Results:
- Statistically optimized feature selection improved model performance.
- XGBoost model achieved an F1-score of 0.916 and AUC-ROC of 0.895.
- Key predictors included infertility, dysmenorrhea, pain levels, and cyclic intestinal pain.
Conclusions:
- Machine-learning models show potential for non-invasive endometriosis risk triage.
- The study utilized a clinically verified cohort and leakage-aware validation.
- External validation and prospective assessment are necessary for clinical implementation.
