# Comparison of Standard and Free-Hand Region-of-Interest Measurements in Strain Elastography for Differentiating Benign and Malignant Breast Masses

## TL;DR

This study compares different methods of measuring stiffness in breast masses using strain elastography to determine which is better at identifying cancerous versus benign tumors.

## Contribution

The study introduces a direct comparison between standard and free-hand ROI measurements in strain elastography for breast cancer diagnosis.

## Key findings

- Circular SR showed higher diagnostic accuracy (AUC 0.843) compared to all SR (AUC 0.822).
- Both circular and all SR measurements were significantly higher in malignant masses.
- The Itoh score was a highly significant predictor of malignancy.

## Abstract

Introduction

Breast cancer remains a prevalent malignancy globally, necessitating accurate and early diagnostic modalities for improved patient outcomes. Strain elastography (SE) is the first and most frequently used elastographic method developed for breast tumor evaluation. SE is widely utilized due to its real-time capabilities and its ability to provide qualitative and semi-quantitative stiffness information.

Objective

This study aimed to compare the diagnostic efficacy of standard region-of-interest (ROI) strain ratio (SR) measurements (circular SR), qualitative Itoh score, and free-hand ROI (all SR) measurements in elastography for distinguishing between benign and malignant breast masses.

Methods

A total of 107 breast masses (55 benign, 52 malignant), all histologically confirmed, were analyzed. Measurements included circular SR (standardized), all SR (operator-defined, free-hand), and the Itoh score (qualitative). Receiver operating characteristic (ROC) curve analysis was used to compare diagnostic performances. Demographic and lesion parameters (age, size, and depth) were analyzed using Mann-Whitney U and chi-square tests.

Results

Both circular SR and all SR measurements were significantly higher in malignant masses (p < 0.0001). Circular SR showed a marginally superior overall diagnostic accuracy with an area under the curve (AUC) of 0.843 (sensitivity: 86.5%, specificity: 76.4%, cut-off > 3.30). The all SR had an AUC of 0.822 (sensitivity: 88.5%, specificity: 74.5%, cut-off > 3.20). The qualitative Itoh score was also a highly significant predictor of malignancy (p< 0.0001). Patients with malignant masses were significantly older (p = 0.0008), but lesion size (p = 0.0596) and depth (p = 0.6090) were not significant differentiating factors.

Conclusion

The standard ROI (circular SR) provides a slightly higher overall diagnostic accuracy, primarily driven by improved specificity. However, the free-hand ROI (all SR) offers higher sensitivity, which is critical in cancer screening. Both quantitative methods, alongside the qualitative Itoh score, are powerful tools for enhancing the diagnostic confidence in breast lesion characterization.