Predictive value of elastography (Shear-Wave and Elasticity Index) in differentiating benign from malignant breast lesions: a retrospective study in Peru

Variable	n (189)	%/Median (IQR)
Age (years)	58 (48-68)	—
Family history of cancer
Yes	42	22.22
No	140	74.07
Not declared	7	3.70
Lesion laterality
Right	83	43.90
Center	106	56.10
Previous mammography	189
Yes	134	70.90
No	55	29.10
Ultrasound BI-RADS category
4A	21	11.10
4B	70	37.00
4C	16	8.50
5	82	43.40
Ultrasound tumor size (mm)	20 (17-28)	—
Lesion depth	16 (12-20)
Elasticity index	2.57 (2-3)	—
Shear-Wave Elastography (SWE, kPa)	54 (49-59)	—
Histopathological diagnosis	85
Malignant	54	63.50
Benign	31	36.50
Breast cancer subtype	n = 35
Luminal A	7	20.00
Luminal B	22	62.90
HER2-positive	4	11.40
Triple negative	2	5.70
Histological grade	2 (2-3)	—

1 Data are expressed as frequencies (%) or medians (interquartile ranges, IQRs), as appropriate.

1 BI-RADS = Breast Imaging Reporting and Data System; SWE = shear wave elastography

TABLE 2.

Spearman’s Rho correlation between lesion characteristics and elastography measurements

Rho de Spearman			Elasticity index	SWE (kPa)
Lesion size (mm)	Spearman’s ρ	1.000	0.226	0.250
	p - value		0.002	0.001
	N	189	188	189
Lesion depth	Spearman’s ρ	1.000	-0.037	0.102
	p - value		0.622	0.176
	N	177	176	177

1 Spearman’s rho correlation coefficients are shown for the associations between lesion size or lesion depth and elastography parameters. Values of p < 0.05 were considered statistically significant. The autocorrelation values (ρ = 1.000) correspond to the variable with itself.

1 SWE = shear wave elastography

TABLE 3.

Comparison of elastography values according to histopathological diagnosis

Histopathological Diagnosis
	Benign (n = 31) Median; [Q1-Q3]	Malignant (n = 54) Median; [Q1-Q3]	p value
Elasticity index	1.9; [1.4–2.5]	2.5; [2–2.8]	0.001
Shear-Wave Elastography (SWE, kPa)	43; [36–49]	53; [48–59]	0.000

1 Comparisons were performed via the Mann–Whitney U test (p < 0.05 was considered significant).

1 Me = median; Q1 = first quartile; Q3 = third quartile.

TABLE 4.

Relationship between BI-RADS category and elastography measurements

Ultrasound BI-RADS	SWE (kPa) Median (IQR)	p value*	Elasticity index Median (IQR)	p value*
4A	38; [29.5–39.5]	0.001	1.6; [1.2–2.05]	0.001
4B	45.5; [39–50]		1.9; [1.5–2.20]
4C	48; [38–55]		2.55; [2–2.80]
5	59; [49–60]		2.7; [2.15–2.80]

1 BI-RADS = Breast Imaging Reporting and Data System; IQR = interquartile range; Me = median; SWE = shear wave elastography.

1* p value obtained via the Kruskal-Wallis test (p < 0.05 considered significant)

Representative shear-wave elastography image of a BI-RADS 4B breast lesion.
Grayscale ultrasound image (left) and corresponding shear-wave elastography map (right) of a BI-RADS 4B breast lesion. The region of interest (ROI) is manually traced along the lesion margins on B-mode imaging. The elastography map shows intermediate stiffness with focal areas of increased rigidity surrounding the lesion, depicted by green to yellow colour coding, without the marked high-stiffness dominance seen in higher-suspicion categories. Quantitative measurements demonstrate a moderate mean SWE value (Emean ≈ 35.5 kPa), illustrating the role of SWE in refining risk stratification within intermediate-suspicion BI-RADS subcategories. Individual SWE values may vary within BI-RADS subcategories; this representative BI-RADS 4B case illustrates moderate mean stiffness with focal areas of increased rigidity, consistent with the heterogeneous nature of this category.

Representative shear-wave elastography image of a BI-RADS 4C breast lesion.
Grayscale ultrasound (left) and corresponding shear-wave elastography (right) of a BI-RADS 4C breast lesion. The region of interest (ROI) is manually delineated along the lesion margins. The elastography map demonstrates marked stiffness heterogeneity with focal areas of increased rigidity (red–orange colour scale). Quantitative measurements show an elevated mean SWE value (Emean ≈ 49.9 kPa) and high maximum stiffness, findings consistent with a high suspicion for malignancy and illustrating the added diagnostic value of SWE in advanced BI-RADS subcategories.

TABLE 5.

Elastography measurements according to BI-RADS 4A and 4B subcategories

Parameter	BI-RADS 4a Median [Q1-Q3]	BI-RADS 4b Median [Q1-Q3]	p value*
Shear-Wave Elastography (SWE), kPa	38.0 [29.5–39.5]	45.0 [39.0–50.0]	< 0.001
Elastography Index	1.60 [1.20–2.05]	1.90 [1.50–2.20]	0.030

1 Q1 = first quartile; Q3 = third quartile;

1* = Statistical test: Mann–Whitney U test

TABLE 6.

Spearman correlation between tumor size and elastography parameters

Parameter	Spearman’s ρ	p value	R² (%)
Elastography Index	0.226	0.002	5.1
Shear-Wave Elastography (SWE, kPa)	0.250	0.001	6.3

1 Spearman’s rank correlation test. R² represents the percentage of variability explained by tumor size. p < 0.05 was considered statistically significant.

TABLE 7.

Binary logistic regression for predictors of malignancy

Parameter	B	p value	ORc (95% CI)	p value	ORa (95% CI)
Age (years)	0.110	< 0.001	1.117 (1.060–1.176)	0.004	1.108 (1.033–1.187)
Tumor size (mm)	–0.025	0.074	0.976 (0.949–1.002)	0.122	0.967 (0.928–1.009)
Shear-Wave left	0.102	< 0.001	1.107 (1.047–1.171)	0.793	0.846 (0.242–2.958)
Elastography Index	1.278	0.001	3.588 (1.648–7.813)	0.106	1.085 (0.983–1.198)
Ultrasound BI-RADS	1.459	< 0.001	4.304 (2.411–7.683)	0.007	2.932 (1.343–6.400)

1 Binary logistic regression model. p < 0.05 was considered statistically significant

1 CI = confidence interval; ORa = adjusted odds ratio obtained from the multivariate logistic regression model; ORc = crude odds ratio (unadjusted).

TABLE 8.

Binary logistic regression model using clinically selected variables

Parameter	B	p value	Adjusted OR (ORa)	95% CI (Lower)	95% CI (Upper)
Age (years)	0.110	< 0.001	1.117	1.054	1.783
Elastography Index	0.422	0.435	1.524	0.530	4.389
SWE (kPa)	0.094	0.016	1.099	1.017	1.186

1 CI = confidence interval; ORa = adjusted odds ratio; p < 0.05 was considered statistically significant

TABLE 9.

Binary logistic regression model excluding age

Parameter	B	p value	Adjusted OR (ORa)	95% CI (Lower)	95% CI (Upper)
Elastography Index	0.671	0.117	1.955	0.846	4.521
SWE (kPa)	0.079	0.011	1.082	1.018	1.150

1 CI = confidence interval; ORa = adjusted odds ratio; SWE = shear wave elastography

1 p < 0.05 was considered statistically significant

TABLE 10.

Area under the ROC curve (AUC) for elastography parameters

Parameter	AUC	SE	p value*	95% CI (Lower-Upper)
Elastog raphy Index	0.725	0.059	0.001	0.610-0.840
SWE (kPa)	0.768	0.056	< 0.001	0.659-0.877

1 p values were derived from the DeLong test. Statistical significance was defined as p < 0.05.

1 SWE = shear wave elastography

TABLE 11.

Bootstrap validation of optimal cut-off points for elastography parameters

Method	Original cut-off	Bootstrap mean	Bootstrap 95% CI	Coefficient of Variation
SWE (kPa)	47.5	47.49	40.5-51.5	4.7%
Elastography Index	2.25	2.054	1.6-2.367	13.0%

1 Cut-off points were derived from the Youden index and validated via 1,000 bootstrap iterations. Lower coefficients of variation indicate more stable and reproducible thresholds.

1 SWE = shear wave elastography

TABLE 12.

Diagnostic performance of elastography parameters

SWE	Diagnostic		Sensitivity (%)	Specificity (%)	PPV(%)	N PV(%)
SWE	Malignant (n = 54)	Benign (n = 31)	Sensitivity (%)	Specificity (%)	PPV(%)	N PV(%)
≥ 47.50 kPa	43	10	79.6	67.7	81.1	65.6
< 47.50 kPa	11	21	79.6	67.7	81.1	65.6
Elasticity index
≥ 2.25	36	9	66.7	71.0	80.0	55.0
< 2.25	18	22
Total	54	31

1 PPV = positive predictive value; NPV = negative predictive value; SWE = shear wave elastography

1 The cut-off values were determined via the Youden index.

Receiver operating characteristic (ROC) curves for shear wave elastography (SWE) and the elastography index.
ROC curves comparing the diagnostic performance of SWE (blue line) and the elastography index (red line) for distinguishing malignant from benign breast lesions. The diagonal dashed line represents the reference line of no discrimination. Both parameters demonstrated discriminative ability above chance, with SWE showing a greater area under the curve (AUC = 0.768) than the elastography index (AUC = 0.725), indicating superior overall diagnostic performance.

References

Authors

Metrics

Articles in this issue

DOI: https://doi.org/10.2478/raon-2026-0018 | Journal eISSN: 1581-3207 | Journal ISSN: 1318-2099

Journal RSS Feed

Language: English

Page range: 204 - 216

Submitted on: Dec 17, 2025

Accepted on: Feb 4, 2026

Published on: Apr 16, 2026

Published by: Association of Radiology and Oncology

In partnership with: Paradigm Publishing Services

Publication frequency: 4 issues per year

Keywords:

breast elastography,

shear wave elastography,

elasticity index,

BI-RADS,

breast neoplasms,

diagnostic ultrasonography,

Related subjects:

© 2026 Gaston (Wilmer) Mendoza, Nancy Muñoz, Jorge Aguilar, Jorge Ayón, Diana Díaz-Llontop, Samantha Mendoza, Claudia Quiñones, Richard Castillo, Luis Taxa, published by Association of Radiology and Oncology
This work is licensed under the Creative Commons Attribution 4.0 License.

Volume 60 (2026): Issue 2 (June 2026)

Predictive value of elastography (Shear-Wave and Elasticity Index) in differentiating benign from malignant breast lesions: a retrospective study in Peru

Figures & Tables

TABLE 1.

TABLE 2.

TABLE 3.

TABLE 4.

FIGURE 1.

FIGURE 2.