# ECO-SCORE: Development of a New Ultrasound Score for the Study of Cystic and Solid-Cystic Adnexal Masses Based on Imaging Characteristics

## TL;DR

This paper introduces ECO-SCORE, a new ultrasound-based tool for diagnosing adnexal masses that outperforms existing models by relying only on imaging features.

## Contribution

ECO-SCORE is a novel ultrasound score that improves diagnostic accuracy by using only imaging characteristics, avoiding demographic or tumor-marker data.

## Key findings

- ECO-SCORE achieved an AUC of 97.08%, outperforming existing models like ADNEX and IOTA.
- Key predictors of malignancy included irregular contour and vascularization within solid areas.
- The model showed high sensitivity (92.98%) and specificity (88.88%) in diagnosing adnexal masses.

## Abstract

The accurate diagnosis of adnexal masses is a critical challenge in gynecological practice. Current ultrasound-based models, such as the ADNEX model, IOTA Simple Rules, and O-RADS, have demonstrated good diagnostic performance but are limited by the inclusion of demographic factors and solid confounding lesions. This study aimed to develop and validate a novel ultrasound score (ECO-SCORE) for cystic and solid-cystic lesions based solely on imaging characteristics to improve diagnostic accuracy and applicability in clinical practice. Methods: We conducted a retrospective study on 330 women diagnosed with adnexal masses, including 251 benign and 79 malignant cases. Ultrasound features were analyzed using logistic regression to identify key predictors of malignancy. A new scoring model was developed, excluding demographic or tumor-marker data. Diagnostic performance metrics, including sensitivity, specificity, AUC, and odds ratios, were calculated and compared to existing models using a testing set (20% of the data). Results: The ECO-SCORE achieved an AUC of 97.08%, outperforming ADNEX model (87.5%), IOTA Simple Rules (85.7%), and O-RADS (87.5%). Sensitivity and specificity were 92.98% and 88.88%, respectively, with an odds ratio of 106. Key predictors included irregular contour, absence of acoustic shadows, vascularization within solid areas, and vascularization of papillae. Conclusions: The ECO-SCORE demonstrated superior diagnostic accuracy compared to established models, highlighting its potential as a reliable tool for assessing adnexal masses using ultrasound features exclusively. Further multicenter validation is needed to confirm its robustness across different clinical settings.