Evaluating diagnostic significance: The utilization of elastography and contrast-enhanced ultrasound for differential diagnosis in breast lesions

Abstract

OBJECTIVE:

The primary aim of this study is to assess the diagnostic efficacy of elastography and contrast-enhanced ultrasound (CEUS) in the identification of breast lesions subsequent to the optimization and correction of the BI-RADS category 4 classification obtained through conventional ultrasound. The objective is to augment both the specificity and accuracy of breast lesion diagnosis, thereby establishing a reliable framework for reducing unnecessary biopsies in clinical settings.

METHODS:

A cohort comprising 50 cases of breast lesions classified under BI-RADS category 4 was collected during the period from November 2022 and November 2023. These cases were examined utilizing strain elastography (SE), shear wave elastography (SWE), and CEUS. Novel scoring methodologies for ultrasonic elastography (UE) and CEUS were formulated for this investigation. Subsequently, the developed UE and CEUS scoring systems were used to refine and optimize the conventional BI-RADS classification, either in isolation or in conjunction. Based on the revised classification, the benign group was classified as category 3 and the suspected malignant group was classified as category 4a and above, with pathological results serving as the definitive reference standard. The diagnostic efficacy of the optimized UE and CEUS, both independently and in combination, was meticulously scrutinized and compared using receiver operating characteristic (ROC) curve analysis, with pathological findings as the reference standard.

RESULTS:

Within the study group, malignancy manifested in 11 cases. Prior to the implementation of the optimization criteria, 78% (39 out of 50) of patients underwent biopsies deemed unnecessary. Following the application of optimization criteria, specifically a threshold of≥8.5 points for the UE scoring method and≥6.5 points for the CEUS scoring method, the incidence of unnecessary biopsies diminished significantly. Reduction rates were observed at 53.8% (21 out of 39) with the UE protocol, 56.4% (22 out of 39) with the CEUS protocol, and 89.7% (35 out of 39) with the combined UE and CEUS optimization protocols.

CONCLUSION:

The diagnostic efficacy of conventional ultrasound BI-RADS category 4 classification for breast lesions is enhanced following optimized correction using UE and CEUS, either independently or in conjunction. The application of the combined protocol demonstrates a notable reduction in the incidence of unnecessary biopsies.

Keywords

BI-RADS 4 categories breast lesions contrast-enhanced ultrasound elastography optimization correction scoring method

1 Introduction

As per the 2020 statistics released by GLOBOCAN, female breast cancer has surpassed lung cancer to become the most prevalent malignant tumor and the fifth leading cause of cancer-related fatalities worldwide. Breast cancer constitutes approximately 24.5% of all reported cases and contributes to 15.5% of cancer-related deaths in the global female population [1]. Currently, the predominant diagnostic framework used in clinical practice for breast lesion diagnosis through routine ultrasound is the fifth edition of the Breast Imaging Reporting and Data System (BI-RADS), published in accordance with the 2013 update by the American College of Radiology (ACR) [2]. However, the interpretative scope of images derived from a single conventional ultrasound (CUS) examination primarily encompasses morphological analysis, thus limiting the provision of detailed imaging information on breast lesions. In the diagnostic context of CUS, recommendations for BI-RADS category 4 or higher often require a biopsy, posing a substantial risk of identifying malignancies across a broad spectrum and potentially contributing to an increase in unnecessary biopsies.

Ultrasound elastic imaging technique can evaluate the hardness information of breast lesions noninvasively and objectively, while hard tissues are easily considered as malignant lesions. Therefore, it is a new diagnostic idea to use elastic imaging technique to diagnose benign and malignant breast lesions.

Ultrasonic elastic imaging can be divided into shear wave elastic imaging and strain elastic imaging. Strain elastic imaging is a relative method to compare the external force changes of the lesion with the surrounding normal tissue, and shear wave elastic imaging is a quantitative method to quantify the hardness of the lesion by measuring the shear wave propagation speed [3]. EFSUMB guidelines indicate that strain-gauge elastography and shear-wave elastography can complement conventional ultrasound to improve the sensitivity and specificity of BI-RADS [4]. Guidelines set forth by the World Federation of Ultrasound in Medicine and Biology affirm its capacity not only to discern between benign and malignant breast lesions, encompassing both nodular and non-nodular varieties, but also to assess the efficacy of neoadjuvant chemotherapy [5]. Georgieva et al. showed that the sensitivity and specificity of strain elastic imaging combined with conventional ultrasound in the diagnosis of breast lesions were 97% and 82%, while the sensitivity and specificity of conventional ultrasound alone were 92% and 73% [6]. Yu et al. showed that the diagnostic efficiency of shear wave elastic imaging in different systems was similar. The diagnostic efficiency was the best when the critical value of Emax in LOGIQ E9 system was 174.5 kPa and Esd in Aixplorer US system was 12.7 kPa. Their sensitivity and specificity were 100.0%, 80.3%, 94.4%, 80.3%, respectively [7]. Cantisani et al. evaluated the diagnostic performance of the qualitative and quantitative information of SE and SWE techniques in the identification of benign and malignant breast lesions, and the results showed that the combination of conventional BI-RADS with SE and SWE isan effective way to improve the classification evaluation of BI-RADS, and may be helpful in the identification of benign and malignant breast lesions [8]. Golatta et al. explored the potential value of SWE and SE in breast cancer diagnosis, and conducted an international multicenter clinical trial in 1288 women with BI-RADS 3 to 4c breast lesions. The results showed that combining SWE and SE into routine ultrasound diagnosis to reclassify BI-RADS 4a patients. It can reduce unnecessary biopsies [9]. Numerous studies have demonstrated that additional quantitative parameters from shear wave elastography (SWE), such as the stiffness gradient and anisotropy factor, offer diagnostic insights from diverse perspectives and exhibit significant value in distinguishing between benign and malignant breast lesions [10 –13]. With regard to contrast-enhanced ultrasound (CEUS) technology, the usage of microbubble contrast agents enables real-time imaging of minute, low-velocity blood flow, imperceptible by CUS alone, providing information on microcirculatory perfusion within the lesion [14]. Prior research has indicated that abnormal angiogenesis is pivotal for the growth and metastasis of malignant tumors [15, 16]. Breast cancer tissues produce considerable amounts of vascular growth factors inducing abnormal angiogenesis. Consequently, CEUS, in theory, can non-invasively assess the degree of angiogenesis in lesions, thereby assisting in the differential diagnosis of benign and malignant aspects of lesions. EFSUMB guidelines state that a large number of studies have shown that contrast-enhanced ultrasound has achieved good initial results in the differential diagnosis of breast lesions [17, 18]. Janu et al. showed that the qualitative and quantitative analysis of CEUS helped to distinguish between type 3 and type 4 lesions, thereby reducing unnecessary biopsies for benign lesions [19]. Zhang et al. showed that CEUS can improve the specificity of diagnosis of malignant non-mass lesions, and the combined application of CEUS and conventional BI-RADS has high sensitivity and specificity for their differential diagnosis [20]. Zeng et al. showed that the diagnostic accuracy of conventional ultrasound combined with contrast-enhanced ultrasound in the differential diagnosis of breast granulomatous lobular mastitis (GLM) and breast infiltrating ductal carcinoma (IDC) was 97.45%, and the AUC was 0.965 [21]. In addition, contrast-enhanced ultrasound has a good diagnostic advantage for small breast lesions. Yu et al. suggested that CEUS could help improve the diagnostic accuracy of BI-RADS 4a masses less than 2 cm in diameter, with diagnostic sensitivity and specificity of 90% and 86%, respectively, and an AUC of 0.904 [22]. Xiao et al. studied the application of CEUS in the differential diagnosis of breast lesions < 10 mm, which can improve the specificity while maintaining high sensitivity, and can improve the diagnostic efficiency when combined with conventional BI-RADS [23].

Currently, limited research has been conducted on the diagnostic value of concurrently employing strain elastography (SE), sound touch elastography (STE), and CEUS. The primary objective of this study is to assess the diagnostic effectiveness of UE and CEUS subsequent to the refinement and adjustment of the BI-RADS categorization derived from CUS for breast lesions. The goal is to augment the specificity and accuracy of breast lesion diagnosis, ultimately providing a reliable basis for the reduction of unnecessary biopsies in clinical practice.

2 Data and methods

2.1 Clinical data

From November 2022 to November 2023, we conducted a study at Jilin Province People’s Hospital, involving the examination of 50 cases of breast lesions in 42 female patients. The Ethics Committee of Jilin Province People’s Hospital (NO.2023146) reviewed and approved the study, and signed informed consent were provided by all patients. Inclusion criteria: ding172; BI-RADS category 4 diagnosed via CUS. ding173; All participants underwent CEUS examination voluntarily. ding174; The lesions had not undergone prior treatment. ding175; Pathological examination results were available. Exclusion criteria: ding172; Pregnant or lactating women. ding173; Post-implantation of breast prosthesis. ding174; Presence of concurrent serious underlying diseases or a history of other malignant tumors. ding175; The size of the lesion exceeded the coverage capacity of CEUS and UE images.

2.2 Instruments and study methods

2.2.1 Examination instruments and reagents

In this study, the implementation of CEUS, CUS, and UE was carried out using the Mindray Resona ultrasonic diagnostic instrument. A high-frequency line array probe (model L14-5WU) was chosen, and the frequency of the probe was configured to range between 5 and 14 MHz. The ultrasound contrast agent employed was Sonovue, consisting of 59 mg sulfur hexafluoride.

2.2.2 Study methods

2.2.2.1 Image collection and analysis Routine ultrasound examinations were conducted by two physicians with over 10 years of experience in breast ultrasound diagnosis, adhering to the 2013 version of the American College of Radiology (ACR) BI-RADS for risk stratification. In cases of classification discrepancies, the physicians engaged in discussions and reached a consensus.

During UE examinations, the focal point was set on the largest diameter section of the lesion at the center of the screen. The sampling frame, encompassing normal tissues, was 2 to 3 times the size of the lesion. Patients were instructed to breathe calmly, and three frames of SE and STE elastic static images meeting standard requirements were selected. This involved ensuring that the pressure cue sensing position of the SE map was at a medium-low level, the motion stability of the STE map was rated 5 stars, and the confidence of the quality control map was≥90%. Image analysis and measurement of each parameter were conducted separately, and the average value was recorded.

Qualitative analysis: The assignment index was determined using the strain-modified five-point method as proposed by Luo et al. [24] 1 point: the lesion is mostly green in color and encompasses the complete or a significant section of the area; 2 points: the lesion is predominantly red in color with a green border; 3 points: the lesion area exhibits similar proportions of red and green; 4 points: a lesion with a primarily red appearance and minor green infiltration; 5 points: the lesion and its adjacent tissues have a crimson hue, irrespective of the presence or absence of green pigmentation within.

Quantitative analysis: The SE of the mass was selected using the diagnostic tool. The fat-to-lesion strain ratio (FLR) was calculated as follows: lesion A was outlined using the tracing approach, while a certain range of normal adipose tissue B₂ (blue region) was highlighted to the greatest extent possible at the same depth as the lesion; FLR = B/A. The shell dimensions were adjusted by 2 mm in accordance with the elasticity and elasticity ratio of the mass in the diagnostic instrument as determined by the STE. With the use of the tracing method, the greatest diameter portion A₁ and the orthogonal section A₂ of the lesion were delineated. Simultaneously, an attempt was made to delineate a specific range of healthy glandular tissues B₁ and B₂ to an extent equivalent to that of A₁ and A₂. The parameters obtained from the post-processing of the diagnostic ultrasonography equipment were as follows: the total area of the lesion, denoted as A1’ (Lesion + Shell), for each value of the elastic modulus (Emax, Emean, Emin, Esd, Eratio). Furthermore, the following parameters were calculated: ding172; Stiffness gradient SG[10]: (A₁Emax–A₁Emean) + A₁Emax; ding173; Anisotropy factor AF[12]: (E_{lesion maximal section}-E➁), which resulted in the AFmax, AFmean, AFsd, and AFratio of the overall region of the lesion, A1’. Employing pathological results as the benchmark, we conducted a statistical analysis of each parameter to discern variances between benign and malignant lesions. Parameters exhibiting significant differences were chosen for the construction of receiver operating characteristic (ROC) curves, facilitating the determination of the most effective diagnostic cut-off value for distinguishing benign and malignant breast lesions (≥cut-off value indicative of malignancy). Assigning a score of 0 to benign indications and a score of 1 to malignant indications, the SE and STE scores were subsequently integrated into the UE scoring system. This algorithm was then applied to assign scores to the study participants, allowing for the testing of statistical differences and the construction of the ROC curve to pinpoint the optimal diagnostic cut-off value.

During the CEUS examination, the focal point was set on the largest diameter section of the lesion at the center of the screen. Subsequently, the CEUS mode was initiated with a mechanical index (MI) of 0.077. Following this, a bolus injection of 4.8 mL of sulfur hexafluoride microbubble suspension and 10 mL of saline was administered into the left elbow vein of the patient to flush the tubes. After the completion of the bolus injection, more than 100 seconds of video data were recorded. The recorded video data were then replayed to qualitatively characterize the CEUS, and the region of interest (ROI) was delineated, with ROI₁ covering the lesion and ROI₂ encompassing the normal glandular tissue surrounding the lesion. Subsequently, analysis software was used to fit and generate the time-signal intensity curve (TIC), allowing for the acquisition and recording of each quantitative parameter related to the lesion and the normal glandular tissue.

Qualitative characteristics: The assignment of index values took into account the following qualitative features: enhancement intensity (equal, low, no/high), enhancement pattern (homogeneous/inhomogeneous), perfusion pattern (same, slow, no/fast), enhancement range (non-expansion/expansion), peripheral morphology of the lesion (regular/irregular), and presence of perforator vessels (none/present) [25].

Quantitative analysis: This includes basic intensity (BI), arrival time (AT), time to peak (TTP), peak intensity (PI), ascending slope (AS), half time of peak intensity (DT/2), descending slope (DS), and amplitude (AMP) = PI-BI. Due to the abnormal blood supply proliferation of malignant foci, a new relative ratio parameter (lesion/adjacent normal glands ratio) was added in this research to study PI and AMP. Specifically, PIratio = PI_lesion/PI_{normal glandular tissue}, and AMPratio = AMP_lesion/AMP_{normal glandular tissue}. By employing pathological findings as the benchmark, we examined the statistical disparities among each parameter in the identification of benign and malignant lesions. To determine the best cut-off value for distinguishing benign from malignant breast lesions (≥cut-off value indicating malignancy) and to screen the characteristics to be included in the assignment indexes, ROC curves were constructed using parameters with significant differences. In the assignment index, benign indications were allocated a score of 0, whilst malignant indications were assigned a score of 1. The assignment index scores were integrated into the CEUS scoring mechanism. The ROC curve was generated to ascertain the best diagnostic cutoff value, statistical differences were assessed, and scores were assigned to the study participants using this approach.

2.2.2.2. Optimized correction scheme for conventional BI-RADS Both UE and CEUS independently implement an optimization correction scheme: When the cumulative score is greater than the designated cut-off value, the BI-RADS classification is elevated by one level. Conversely, when the total score greater than or equal to the cut-off value, the BI-RADS classification is reduced by one level. In the joint optimization and correction scheme for UE and CEUS, if both UE and CEUS scores are greater than or equal to the cutoff value, the BI-RADS classification is heightened by two levels. Conversely, if both UE and CEUS scores are less than the cutoff value, the BI-RADS classification is lowered by two levels. In all other cases, the BI-RADS classification remains unaltered. It is important to note that the optimization correction processes cease adjustments downward if the initial classification is category 3 and refrain from adjustments upward if the initial classification is category 5.

2.3 Statistical analysis

The data were subjected to analysis using SPSS 26.0 statistical software. Normality of the measurement data was assessed, and for normally distributed data, a t-test was used, while non-normally distributed data underwent analysis using a rank-sum test. Count data were scrutinized using Fisher’s exact probability method, and the McNemar test was used for rate comparisons. Pathological results served as the gold standard for constructing the ROC curve. The area under the ROC curve (AUC) value was calculated and analyzed, with the optimal diagnostic cut-off value determined through the maximum Youden’s index. The diagnostic efficacy of each statistically significant quantitative parameter and scoring method in identifying benign and malignant breast lesions was thoroughly analyzed. Subsequently, BI-RADS categories≤3 were classified as the benign group, while BI-RADS categories≥4a (with malignant risk > 2%) were categorized as the suspected malignant group. The diagnostic value of UE, CEUS alone optimization scheme, and the combined optimization scheme were meticulously analyzed and compared. A significance level of P < 0.05 was considered indicative of a statistically significant difference.

3 Results

3.1 General data and results of BI-RADS classification of CUS

A total of 50 breast lesions were examined in 42 female patients, with a maximum diameter ranging from 2.4 mm to 66.0 mm and a mean diameter of 13.64±10.76 mm. The ages of the patients varied from 29 to 81 years, with a mean age of 48.70±11.92 years. Pathological findings revealed 39 cases of benign lesions, comprising of 22 cases of fibroadenoma, 10 cases of adenosis, 5 cases of intraductal papilloma, 1 case of tubular adenoma, and 1 case of complex cyst. Additionally, there were 11 cases of malignant lesions, including 9 cases of non-specific invasive carcinoma, 1 case of invasive lobular carcinoma, and 1 case of mucinous carcinoma. The BI-RADS classification of CUS with pathological control indicated that among the lesions in category 4a, 22 cases were benign. In category 4b, there were 19 cases, with 17 being benign and 2 malignant. The lesions in category 4c comprised 9 cases, all of which were malignant.

3.2 Results of each quantitative parameter analysis and the diagnostic efficacy of UE and CEUS

The examination of each quantitative parameter demonstrated that the FLR, gland-to-lesion strain ratio (GLR), A₁’ (Emax, Emean, Esd, Eratio), SG, and A’AFratio in malignant lesions were all notably higher than those observed in benign lesions, as indicated in Table 1 (P < 0.05). The ROC curve was constructed using the pathological results as the gold standard. Table 2 presents the optimal diagnostic efficacy of each quantitative parameter when the Youden’s index was chosen as the maximum.

Table 1
Results of comparative analyses of quantitative parameters of UE

Quantitative parameter Pathological examination results P value

Benign cases Malignant cases

SE parameter

FLR 2.88 (2.37,3.79) 4.74±1.77 0.002^*

STE parameter

A₁’Emax 53.01 (36.75,75.90) 109.84±36.23 0.000^*

A₁’Emean 22.32 (15.97,32.25) 35.26±8.04 0.006^*

A₁’Emin 6.27 (5.46,10.71) 8.57±3.81 0.419

A₁’Esd 9.06 (5.37,13.15) 13.78±4.52 0.020^*

A₁’Eratio 1.18 (1.01,1.38) 2.48±1.39 0.001^*

SG 70.84 (46.21,107.73) 134.77±60.95 0.004^*

A’AFmax 140.42 (15.29,476.11) 657.92(5.02,4853.91) 0.153

A’AFmean 9.06 (1.93,108.37) 35.40(4.93,46.51) 0.566

A’AFsd 3.31 (0.24,18.40) 5.76(0.55,25.20) 0.699

A’AFratio 0.04 (0.00,0.22) 0.53(0.10,3.28) 0.005^*

Quantitative parameter	Pathological examination results	P value
SE parameter
FLR	2.88 (2.37,3.79)	4.74±1.77	0.002^*
STE parameter
A₁’Emax	53.01 (36.75,75.90)	109.84±36.23	0.000^*
A₁’Emean	22.32 (15.97,32.25)	35.26±8.04	0.006^*
A₁’Emin	6.27 (5.46,10.71)	8.57±3.81	0.419
A₁’Esd	9.06 (5.37,13.15)	13.78±4.52	0.020^*
A₁’Eratio	1.18 (1.01,1.38)	2.48±1.39	0.001^*
SG	70.84 (46.21,107.73)	134.77±60.95	0.004^*
A’AFmax	140.42 (15.29,476.11)	657.92(5.02,4853.91)	0.153
A’AFmean	9.06 (1.93,108.37)	35.40(4.93,46.51)	0.566
A’AFsd	3.31 (0.24,18.40)	5.76(0.55,25.20)	0.699
A’AFratio	0.04 (0.00,0.22)	0.53(0.10,3.28)	0.005^*

Note: Normal distribution data: X±S; Non-normal distribution data: M(Q1,Q2). ^* indicates statistically significant difference, P < 0.05.

Table 2

Diagnostic effectiveness of statistically significant UE parameters

Parameter	Cut-off	Se %	Sp %	ACC %	PV+%	PV–%	LR+	LR–	AUC
SE parameter
FLR	3.96	81.8	82.1	82.0	56.3	94.1	4.57	0.22	0.803
STE parameter
A₁’Emax	78.89	81.8	82.1	82.0	56.3	94.1	4.56	0.22	0.872
A₁’Emean	27.74	90.9	64.1	70.0	41.7	96.2	2.53	0.14	0.774
A₁’Esd	10.10	81.8	61.5	66.0	37.5	92.3	2.13	0.30	0.732
A₁’Eratio	1.39	90.9	76.9	80.0	52.6	96.8	3.94	0.12	0.829
SG	96.59	72.7	74.4	74.0	44.4	90.6	2.84	0.37	0.783
A’AFratio	0.06	90.9	56.4	64.0	37.0	95.7	2.09	0.16	0.779

Note: Cut-off: cut-off value; Se: sensitivity; Sp: specificity; ACC: accuracy; PV+: positive predictive value; PV–: negative predictive value; LR+: positive likelihood ratio; LR–: negative likelihood ratio.

3.3 Results of the quantitative parameter analysis and the diagnostic efficacy of CEUS

The quantitative analysis of TIC parameters revealed that the DT/2, AMP, and PIratio in malignant lesions exhibited higher values compared to benign lesions, with statistically significant differences (P < 0.05), as presented in Table 3. Employing pathological results as the gold standard, the ROC curve was generated. Table 4 delineates the diagnostic performance of each quantitative parameter when the Youden index was chosen as the maximum.

Table 3
Results of comparative analyses of quantitative parameters of CEUS

CEUS parameter Pathological examination results P value

Benign cases Malignant cases

BI 10.82 (7.45,14.81) 7.33 (6.59,9.08) 0.081

AT 6.20 (0.50,11.20) 10.50 (7.60,11.70) 0.246

TTP 18.26±7.39 23.04±7.11 0.062

PI 25.03±8.35 27.59±6.21 0.352

AS 0.70±0.42 0.88±0.40 0.205

DT/2 58.62±21.11 74.06±21.00 0.037^*

DS 0.15(0.10,0.21) 0.20±0.10 0.392

AMP 12.89±7.43 18.65±8.29 0.032^*

PIratio 0.89 (0.70,1.18) 1.24±0.44 0.043^*

AMPratio 1.43 (0.82,2.34) 2.19±1.13 0.101

CEUS parameter	Pathological examination results	P value
BI	10.82 (7.45,14.81)	7.33 (6.59,9.08)	0.081
AT	6.20 (0.50,11.20)	10.50 (7.60,11.70)	0.246
TTP	18.26±7.39	23.04±7.11	0.062
PI	25.03±8.35	27.59±6.21	0.352
AS	0.70±0.42	0.88±0.40	0.205
DT/2	58.62±21.11	74.06±21.00	0.037^*
DS	0.15(0.10,0.21)	0.20±0.10	0.392
AMP	12.89±7.43	18.65±8.29	0.032^*
PIratio	0.89 (0.70,1.18)	1.24±0.44	0.043^*
AMPratio	1.43 (0.82,2.34)	2.19±1.13	0.101

Note: Normal distribution data: X±S; Non-normal distribution data: M(Q1,Q2). ^* indicates statistically significant difference, P < 0.05.

Table 4

Diagnostic efficacy of statistically significant parameters of CEUS

CEUS parameter	Cut-off	Se %	Sp %	ACC %	PV+%	PV–%	LR+	LR–	AUC
DT/2	56.85	81.8	51.3	58.0	32.1	90.9	1.68	0.35	0.689
AMP	16.98	63.6	74.4	72.0	41.2	87.9	2.48	0.50	0.718
PIratio	1.01	72.7	61.5	64.0	34.8	88.9	1.89	0.44	0.702

3.4 Results of assignment indexes and diagnostic value of each scoring method for UE and CEUS

In summary, the quantitative parameters demonstrating statistical significance in UE and CEUS are considered as screening criteria. Combining the quantitative and qualitative scoring indexes of UE and CEUS led to the establishment of their respective scoring criteria. The UE scoring method comprised eight items: quintile, FLR, A₁’ Emax, A₁’ Emean, A₁’ Esd, A₁’ Eratio, SG, and A’AFratio, with a maximum score of 12 points. While the CEUS scoring method included nine items: enhancement intensity, enhancement pattern, perfusion pattern, enhancement range, lesion peripheral morphology, perforator vessel characteristics, DT/2, AMP, and Piratio, with a maximum score of 9 points.

The analysis of the UE and CEUS scoring methods revealed that the median scores for malignant lesions were significantly higher than those for benign lesions (UE: 11 vs. 4; CEUS: 8 vs. 3), with this difference being statistically significant (P < 0.05). The ROC curves for each scoring method were plotted using pathological results as the gold standard, as shown in Fig. 1. Optimal diagnostic cut-off values for UE and CEUS were 8.5 and 6.5 points, respectively, when the maximum Youden’s index was selected. Correspondingly, the AUC values were 0.931 and 0.958. Notably, the AUC of the CEUS scoring method exceeded that of the UE scoring method. However, based on ROC curve analysis, the results indicated no statistically significant difference in the AUC between UE and CEUS (P > 0.05). Detailed information on diagnostic efficacy is available in Table 5.

Fig. 1

ROC curve of each scoring method.

Table 5

Diagnostic efficacy of the scoring methods

Scoring method	Cut-off	Se %	Sp %	ACC %	PV+%	PV–%	LR+	LR–	AUC
UE scoring method	8.5	81.8	92.3	90.0	75.0	94.7	10.64	0.20	0.931
CEUS scoring method	6.5	81.8	97.4	94.0	90.0	95.0	31.91	0.19	0.958

3.5 Results of the diagnostic value of UE and CEUS and in combination with the optimized correction scheme

The recommendation for tissue biopsy is made for categories above 4 by the ACR BI-RADS classification. Consequently, the categories≤3 of BI-RADS were considered as the benign group, while categories≥4a (malignant risk > 2%) of BI-RADS were classified as the suspected malignant group. The results of the statistical analysis revealed that, following the optimization of UE and CEUS alone and in combination, the differences in the ratio of benign lesions between the benign and suspected malignant groups of the three schemes were all found to be statistically significant (P < 0.05), as indicated in Table 6. The ROC curves for each diagnostic approach were plotted with the pathological results as the gold standard. It was observed that the AUCs for the UE, CEUS alone optimization regimen, and the combined optimization regimen were 0.769, 0.782, and 0.949, respectively. The highest AUC was demonstrated by the combined optimization regimen. a comparison of the AUC between the two regimens based on the ROC curves revealed that the AUC difference between the combined optimization regimen and the UE, CEUS alone regimen was statistically significant (P < 0.05). However, the AUC difference between UE and CEUS alone was not statistically significant (P > 0.05). The results of other assessment indexes of diagnostic efficacy revealed that: ding172; The specificity and accuracy of the combined optimization scheme were the best, while the sensitivity remained unchanged. The differences were statistically significant compared to those of the individually optimized UE and CEUS schemes (P < 0.05). Additionally, the differences between the individually optimized UE and CEUS schemes were not statistically significant (P > 0.05). ding173; The accuracy of all three optimization schemes enhanced significantly when compared to before optimization, with a statistically significant difference (P < 0.05). The diagnostic effectiveness is outlined in Table 7.

Table 6
Analysis of the comparative composition ratio of benign and malignant lesions across groups subsequent to optimal correction

Groups after optimization correction Pathological examination results P value

Benign cases Malignant cases

CUS + UE 0.001^*

Benign group 21 0

Suspected malignant group 18 11

CUS + CEUS 0.001^*

Benign group 22 0

Suspected malignant group 17 11

CUS + UE + CEUS 0.000^*

Benign group 35 0

Suspected malignant group 4 11

Groups after optimization correction	Pathological examination results	P value
CUS + UE			0.001^*
Benign group	21	0
Suspected malignant group	18	11
CUS + CEUS			0.001^*
Benign group	22	0
Suspected malignant group	17	11
CUS + UE + CEUS			0.000^*
Benign group	35	0
Suspected malignant group	4	11

Note: ^*Indicates that the difference in the composition ratio of benign and malignant lesion between groups was statistically significant, P < 0.05.

Table 7

A Comparative analysis of diagnostic indexes in the BI-RADS classification prior to and following correction

Diagnostic indicator	Prior to optimization correction	Following optimization and correction
	CUS	UE	CEUS	UE + CEUS
Se	100.0% (11/11)	100.0% (11/11)	100.0% (11/11)	100.0% (11/11)
Sp	0.0% (0/39)	53.8% (21/39)	56.4% (22/39)	89.7% (35/39)^b
ACC	22.0% (11/50)	64.0% (32/50)^a	66.0% (33/50)^a	92.0% (46/50)^ab
AUC	—	0.769	0.782	0.949^b

Note: Se: sensitivity; Sp: specificity; ACC: accuracy. ^aindicates a statistical difference in comparison with the pre-correction, ^bindicates a statistical difference in comparison with all other correction schemes, P < 0.05.

As per the fifth edition of the BI-RADS classification guidelines revised by the ACR, tissue biopsy was recommended for all benign lesions before optimization correction, resulting in a 100% biopsy rate (39/39) for benign cases. However, following the implementation of each optimization correction scheme, including UE, CEUS alone, and the co-optimization scheme, and while maintaining the same biopsy rate for malignant lesions, both the overall biopsy rate and the biopsy rate for benign lesions witnessed a reduction of 58.0% (29/50), 56.0% (28/50), 30.0% (15/50), and 46.2% (18/39), 43.6% (17/39), 10.3% (4/39), respectively. The reduction in the biopsy rate was more prominent with the combined optimization regimen compared to the UE, CEUS alone optimization regimen, and the difference was deemed statistically significant (P < 0.05).

4 Discussion

Novel ultrasound methodologies, such as UE and CEUS, hold the potential to augment the specificity of CUS in the diagnosis of breast lesions, thereby reducing the necessity for unnecessary tissue biopsies [26].

The strain ratio (SR) in SE serves as a relatively objective semiquantitative parameter for discerning between benign and malignant breast lesions. However, there is currently no standardized criterion for determining the type of tissue to be used as a reference [27 –30]. Relying on the the relatively stable texture and hardness of normal adipose tissue components, and drawing from the assertion of Zhou et al. that the diagnostic efficacy of FLR for breast lesions surpasses that of GLR, FLR was selected as the research parameter for this study [31]. The findings revealed that the diagnostic efficacy was optimal when the cutoff value for FLR was set at 3.96, with corresponding sensitivity, specificity, and AUC values of 81.8%, 82.1%, and 0.803, respectively. In comparison to findings in related literature, the diagnostic effectiveness was found to be comparable to previous studies, with a threshold range of 2.27–4.79, sensitivity of 79–95%, specificity of 74–89%, and an AUC of 0.830–0.949 [31 –34]. The diagnostic efficacy of conventional quantitative parameters in STE for distinguishing between benign and malignant breast lesions has been extensively documented. However, there is a lack of consensus regarding the most effective parameters for diagnostic purposes [8 , 36].

The “Shell” function of the Mindray ultrasound diagnostic instrument possesses the capability to quantify the texture and hardness of the region surrounding a lesion. Malignant breast lesions often involve the infiltration of cancer cells into the surrounding interstitium, leading to proliferation and the promotion of fibroplastic reaction. This process results in an increase in the hardness around the lesion [37]. The “Shell” function is capable of quantifying and assessing the qualitative “hard ring sign,” forming the theoretical basis for its ability to differentiate between benign and malignant breast lesions [37]. Consequently, we selected the overall area A₁’ of the lesion, which encompasses 2 mm of tissue in the periphery, as the index for analysis. The findings of this analysis align with previous studies indicating that the elastic modulus of malignant lesions is higher than that of benign lesions, as reported in literature [38 –40]. Tumor heterogeneity is a prominent feature of malignant tumors, displaying morphological heterogeneity due to the presence of different types of cells and stromal components in various regions [41]. Samani et al. demonstrated that the Young’s modulus of normal and malignant breast specimens differed ex vivo, and breast cancers displayed significant elastic heterogeneity [42]. Based on the assumption that the difference between Emean and Emax is greater in heterogeneous than in homogeneous lesions, the heterogeneity of breast lesions can be reflected in this difference, defined as the stiffness gradient (SG) [10]. Anisotropy refers to the variation in properties within fiber-rich tissues in different directions. In elastography, this is evident as varying imaging features when the probe orientation is changed, resulting in different elastic parameter measurements [13]. The cutoff value for SG in this study was higher than that in the study by Shang et al. [11]. This may be attributed to the tracing method used in this study, outlining the entire lesion as the ROI.

This led to an increased cutoff value in contrast to the ROI of the most challenging region chosen for this study, which exhibited a lower mean value. Contrary to the results of Chen et al., the only statistically significant difference between benign and malignant lesions in this study was the A’AFratio [13]. This phenomenon could be attributed to several factors: the current study used a small sample size, the ROI was chosen as the overall A’ of the lesion, and malignant lesions comprised a smaller proportion of the total lesions. However, the AUC of A’AFratio was comparable to that of Chen et al. (AUC = 0.804), indicating a comparable diagnostic impact.

The quantitative parameters of CEUS may vary based on the ultrasound machine manufacturer, the software used for TIC analysis, and the ROI settings. There is a lack of consensus regarding which parameters provide the most effective diagnostic value for breast lesions [43, 44]. Among the quantitative TIC parameters in this study, DT/2, AMP, and PIratio were observed to be higher in malignant lesions compared to benign lesions, and this difference was statistically significant (P < 0.05). A longer DT/2 duration signifies an extended retention of the contrast agent in the lesion, possibly attributed to the heterogeneous vascular structure of malignant lesions characterized by a tortuous and chaotic arrangement, as opposed to the longer vessels with a more regular morphology found in benign lesions [45]. Consequently, the contouring time for malignant lesions is ultimately longer than that for benign lesions. A larger AMP value represents a greater instantaneous perfusion of the contrast medium in the lesion. The PIratio reflects the difference between the peak intensity of the lesion and that of the surrounding normal glandular tissues. A PIratio > 1 indicates that the lesion is highly perfused compared with the surrounding normal glandular tissues, and vice versa. These two parameters indicate the degree of contrast medium aggregation in the lesion from different perspectives. Malignant lesions exhibited higher values than benign lesions, possibly due to the higher density of microvessels in malignant lesions. Additionally, the disorganized and loosely connected arrangement of endothelial cells in malignant lesions leads to increased microcirculatory permeability, causing a significant amount of contrast medium to enter the tissue interstitial space, presenting a hyperperfusion characteristic [45, 46].

In this study, the analysis of the UE score method revealed that malignant lesions in SE qualitative features were predominantly red in the entire area or red in the lesion and its surrounding area (10/11, 90.9%). Regarding quantitative parameters, FLR (9/11, 81.8%), A₁’Emax (9/11, 81.8%), A₁’Emean (10/11, 90.9%), A₁’Esd (9/11, 81.8%), A₁’Eratio (10/11, 90.9%), SG (8/11, 72.7%), and A’AFratio (10/11, 90.9%) surpassed the diagnostic cutoff. In contrast, SE of benign lesions predominantly exhibited an overall green color of the lesion, a red center with a green periphery, or a similar red and green color within the extent of the lesion (35/39, 89.7%). Quantitative parameters of FLR (32/39, 82.1%), A₁’Emax (32/39, 82.1%), A₁’Emean (25/39, 64.1%), A₁’Esd (24/39, 61.5%), A₁’Eratio (30/39, 76.9%), SG (29/39, 74.4%), and A’AFratio (22/39, 56.4%) were below the diagnostic cutoff values. The analysis of the CEUS scoring method in this study revealed that malignant lesions mostly exhibited a fast-forward perfusion pattern (9/11, 81.8%), hyperenhancement (8/11, 72.7%), irregularities in the surrounding tissues (9/11, 81.8%), enlargement of enhancement (11/11, 100%), and the presence of a perforating vessel (9/11, 81.8%) among the qualitative features. Among the quantitative parameters, DT/2 (9/11, 81.8%), AMP (7/11, 63.6%), and PIratio (8/11, 72.7%) exceeded the diagnostic cutoff. Benign lesions tended to exhibit a fast-forward perfusion pattern (23/39, 59.0%), no or low or equal enhancement (23/39, 59.0%), peripheral tissue regularity (37/39, 94.9%), no widening of enhancement (39/39, 100%), and the absence of perforating vessels (29/39, 74.4%) in the qualitative features. In terms of quantitative parameters, the DT/2 (20/39, 51.3%), AMP (29/39, 74.4%), and PIratio (24/39, 61.5%) were all below the diagnostic cutoff values.

The analysis of the UE alone optimization protocol in this study revealed that the BI-RADS classification was upgraded in 3 out of 39 cases (7.7%) of benign lesions. In 2 cases of fibroadenomas, the higher classification was likely due to the increased hardness of the texture caused by the more pronounced proliferation of the fibrotic component within the lesion, which densified the overall structure [47]. Additionally, 1 case of a complex cyst had an upgraded classification, possibly due to its more superficial location (approximately 4 mm from the skin), which reduced the accuracy of the measured parameters of SE and STE [5]. The BI-RADS classification of 2 cases (2/11, 18.2%) of malignant lesions was downgraded. One instance of non-specific invasive carcinoma may be attributed to the elevated presence of cancer cells and hyaluronic acid in the extracellular matrix, leading to a predominantly inflammatory rather than fibro-proliferative response. This results in a reduction in the hardness of the texture of the lesion due to the high water content [48]. One case of mucinous carcinoma was likely caused by the abundant mucus surrounding the cancer cells, leading to a decrease in tissue hardness [48]. In this study, the analysis of CEUS optimization alone revealed that the BI-RADS classification in 1 case (1/39, 2.6%) of benign lesions was upgraded. The qualitative characteristics of this fibroadenoma demonstrated fast-moving inhomogeneous hyperenhancement and visible perforating blood vessels. Additionally, the quantitative parameters of DT/2, AMP, and PIratio were larger than the truncation value, which may be related to the abundant microvessel density of the fibroadenoma [49]. The BI-RADS classification of malignant lesions was downgraded in 2 cases (2/11, 18.2%). The qualitative features of 1 case of non-specific invasive carcinoma and mucinous carcinoma both exhibited fast-entry inhomogeneous low- and no-enhancement, irregular morphology of the surrounding tissues, and enlarged enhancement. The quantitative parameter of only invasive carcinoma had a DT/2 greater than the cut-off value, which may be attributed to the larger scope of necrosis in the internal part of the invasive carcinoma. The mucinous carcinoma was rich in mucus, causing less contrast agent entry. Limited access can lead to a lower score and inaccurate assessment. Following the combined optimization protocol, the classification of 4 cases (4/39, 10.3%) of benign lesions upgraded by UE or CEUS alone remained unchanged, reflecting that the combined diagnosis reduces the misdiagnosis rate of individual diagnoses. Additionally, 2 cases (2/11, 18.2%) of malignant lesions downgraded by UE or CEUS alone remained unchanged, indicating that the combined diagnosis reduced the leakage rate of separate diagnoses. It is worth noting that there was still 1 case (1/11, 9.1%) of mucinous carcinoma that was downgraded to category 4a after the combined optimization protocol. However, based on the ACR BI-RADS classification and treatment recommendations, tissue biopsy was still recommended to avoid clinical leakage. Therefore, morphological characterization of CUS should not be detached from the clinical diagnosis and treatment of breast lesions.

The study results revealed that the combined optimization protocol led to a decreased overall biopsy rate and benign lesion biopsy rate compared to the individual optimization protocol, all while maintaining the detection rate of malignant lesions. Among the benign cases that avoided biopsy in this study, it was observed that the protocols using UE and CEUS alone resulted in downgrading from category 4a to category 3. In contrast, the combined protocols demonstrated downgrades from category 4a to category 3 in 95.5% of cases (21/22) and from category 4b to category 3 in 82.4% of cases (14/17). This indicates that a comprehensive assessment based on the conventional BI-RADS classification, considering the texture and hardness of the lesion and microcirculatory perfusion, may prove beneficial for category 4b benign lesions. However, further large-sample studies are needed to confirm these findings.

This study has several limitations: ding172; The qualitative analysis of CEUS and SE is subjective, and there may be measurement errors in the quantitative parameters; ding173; The study did not incorporate multimodal joint diagnosis with technologies such as firefly imaging, three-dimensional imaging, and other emerging ultrasound techniques; ding174; The sample size in this single-center study is small, which could introduce potential selection bias. Therefore, further validation is needed with a larger sample size. Increasing the sample size for additional validation remains necessary.

5 Conclusion

Conventional ultrasound BI-RADS category 4 classification for breast lesions, using UE and CEUS alone or in combination with optimized correction (scoring≥8.5 points on the UE scoring method and≥6.5 points on the CEUS scoring method), exhibits enhanced diagnostic efficacy. Specifically, the combined protocol demonstrates a notable reduction in the number of unnecessary biopsies.

Footnotes

Acknowledgments

We would like to acknowledge the hard and dedicated work of all the staff that implemented the intervention and evaluation components of the study.

Conflict of interest

The authors declare that they have no competing interests.

Funding

No external funding received to conduct this study.

References

Sung

, Ferlay

, Siegel

, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2021;71(3):209–49.

Mendelson

, Bohm-Velez

, Berg

, et al. ACR BIRADS Ultrasound. In: ACR BI-RADS Atlas, Breast Imaging Reporting and Data System. Reston. VA: American College of Radiology, 2013.

Barr

. Sonographic breast elastography: a primer. J Ultrasound Med. 2012;31(5):773–83.

Săftoiu

, Gilja

, Sidhu

, et al. The EFSUMB Guidelines and Recommendations for the Clinical Practice ofElastography in Non-Hepatic Applications: Update 2018. Die EFSUMB-Leitlinien und Empfehlungen für dieklinische Praxis der Elastografie bei nichthepatischen Anwendungen: Update 2018. Ultraschall Med 2019;40(4):425–53.

Barr

, Nakashima

, Amy

, et al. WFUMB guidelines and recommendations for clinical use of ultrasound elastography: Part 2: breast. Ultrasound Med Biol. 2015;41(5):1148–60.

Georgieva

, Prantl

, Utpatel

, et al. Diagnostic performance of ultrasound strain elastography for differentiation of malignant breast lesions. Clin Hemorheol Microcirc. 2019;71(2):237–47.

, Zhang

, Yin

, et al. Two-dimensional shear wave elastography with two different systems for the diagnosis of breast lesions. Clin Hemorheol Microcirc. 2022;82(1):53–62.

Cantisani

, David

, Barr

, et al. US-Elastography for Breast Lesion Characterization: Prospective Comparisonof US BIRADS, Strain Elastography and Shear wave Elastography. US-Elastografie zur Charakterisierung vonBrustläsionen: Prospektiver Vergleich von US-BI-RADS, Strain-Elastografie und Scherwellen-Elastografie.Ultraschall Med. 2021;42(5):533–40.

Golatta

, Pfob

, Büsch

, et al. The Potential of Shear Wave Elastography to Reduce Unnecessary Biopsies inBreast Cancer Diagnosis: An International, Diagnostic, Multicenter Trial. Das Potenzial vonScherwellen-Elastografie zur Reduktion unnötiger Biopsien in der Brustkrebsdiagnostik: Eine internationale,diagnostische, multizentrische Studie. Ultraschall Med. 2023;44(2):162–8.

10.

Blank

MAB

, Antaki

. Breast Lesion Elastography Region of Interest Selection and Quantitative Heterogeneity: A Systematic Review and Meta-Analysis. Ultrasound Med Biol. 2017;43(2):387–97.

11.

Shang

, Ruan

, Wang

, et al. Utilizing size-based thresholds of stiffness gradient to reclassify BI-RADS category 3-4b lesions increases diagnostic performance. Clin Radiol. 2019;74(4):306–13.

12.

Skerl

, Vinnicombe

, Thomson

, McLean

, Giannotti

, Evans

. Anisotropy of Solid Breast Lesions in 2D Shear Wave Elastography is an Indicator of Malignancy. Acad Radiol. 2016;23(1):53–61.

13.

Chen

, Gao

, Chang

, Wang

, Zeng

, Chen

. Ultrasound shear wave elastography of breast lesions: correlation of anisotropy with clinical and histopathological findings. Cancer Imaging. 2018;18(1):11.

14.

Sridharan

, Eisenbrey

, Dave

, et al. Quantitative Nonlinear Contrast-Enhanced Ultrasound of the Breast. AJR Am J Roentgenol. 2016;207(2):274–81.

15.

Weidner

, Semple

, Welch

, Folkman

. Tumor angiogenesis and metastasis–correlation in invasive breast carcinoma. N Engl J Med. 1991;324(1):1–8.

16.

Lichtenbeld

, Barendsz-Janson

, van Essen

, Struijker Boudier

, Griffioen

, Hillen

. Angiogenic potential of malignant and non-malignant human breast tissues in an in vivo angiogenesis model. Int J Cancer. 1998;77(3):455–9.

17.

Sidhu

, Cantisani

, Dietrich

, et al.The EFSUMB Guidelines and Recommendations for the Clinical Practice of Contrast-Enhanced Ultrasound (CEUS) in Non-Hepatic Applications: Update 2017 (Long Version).

18.

Die EFSUMB-Leitlinien und Empfehlungen für den klinischen Einsatz des kontrastverstärkten Ultraschalls (CEUS) bei nicht-hepatischen Anwendungen: Update 2017 (Langversion). Ultraschall Med. 2018;39(2)e2–e44.

19.

Janu

, Krikavova

, Little

, et al. Prospective evaluation of contrast-enhanced ultrasound of breast BI-RADS 3-5 lesions. BMC Med Imaging.66. Published Jun. 2020;20(1):66. Published 2020 Jun 17.

20.

Zhang

, Jin

, Li

, et al. The role of contrast-enhanced ultrasound in the diagnosis of malignant non-mass breast lesions and exploration of diagnostic criteria. Br J Radiol 2021;94(1120)20200880.

21.

Zeng

, Yang

, Liu

, et al.The discriminatory diagnostic value of multimodal ultrasound combined with blood cell analysis for granulomatous lobular mastitis and invasive ductal carcinoma of the breast. Clin Hemorheol Microcirc. 2023. doi :10.3233/CH-231999.

22.

, Zhang

, Jiang

, Zhou

. The value of contrast-enhanced ultrasound in the diagnosis of BI-RADS-US 4alesions less than 2 cm in diameter. Clin Hemorheol Microcirc. 2023;83(3)195–205. doi :10.3233/CH-221460.

23.

Xiao

, Jiang

, Wu

, Guan

, Qin

, Luo

. Diagnosis of sub-centimetre breast lesions: combining BI-RADS-US with strain elastography and contrast-enhanced ultrasound-a preliminary study in China. Eur Radiol. 2017;27(6):2443–50.

24.

Luo

, Ou

, Zhi

, Zeng

, Yang

. The value of improved elasticity scores in the differential diagnosis of malignant and benign breast tumours. Journal of Modern Clinical Medical Bioengineering. 2006)(05)396–8.

25.

Jia

, Du

, Shi

, et al. The contrast enhanced ultrasound qualitative diagnosis of benign and malignant in BI-RADS 4 breast nodules with different sizes. Chinese Journal of Ultrasonography. 2020;29(4):343–4.

26.

Eisenbrey

, Dave

, Forsberg

. Recent technological advancements in breast ultrasound. Ultrasonics. 2016;70 183–90.

27.

Zhi

, Xiao

, Yang

, Ou

, Wen

, Luo

. Ultrasonic elastography in breast cancer diagnosis: strain ratio vs 5-point scale. Acad Radiol. 2010;17(10):1227–33.

28.

, Zhong

, Yao

, et al. Ultrasonic Elastography Research Based on a Multicenter Study: Adding Strain Ratio after 5-Point Scoring Evaluation or Not. PLoS One. 2016;11(2):e0148330.

29.

Zhao

, Yao

, Zhou

, et al. Strain Elastography: A Valuable Additional Method to BI-RADS? Strain Elastografie: Eine wertvolle Zusatzmethode für BI-RADS?. Ultraschall Med. 2018;39(5):526–34.

30.

Park

, Kim

, Yun

, et al. Comparison of One- and Two-Region of Interest Strain Elastography Measurements in the Differential Diagnosis of Breast Masses. Korean J Radiol. 2020;21(4):431–41.

31.

Zhou

, Zhou

, Zhan

, Jia

, Dong

, Yang

. Elastography ultrasound for breast lesions: fat-to-lesion strain ratio vs gland-to-lesion strain ratio. Eur Radiol. 2014;24(12):3171–7.

32.

Thomas

, Degenhardt

, Farrokh

, Wojcinski

, Slowinski

, Fischer

. Significant differentiation of focal breast lesions: calculation of strain ratio in breast sonoelastography. Acad Radiol. 2010;17(5):558–63.

33.

Fischer

, Peisker

, Fiedor

, et al. Significant differentiation of focal breast lesions: raw data-based calculation of strain ratio. Ultraschall Med. 2012;33(4):372–9.

34.

Ozsoy

, Acar

, Barca

, et al. Diagnostic performance of real-time strain sonoelastography in BI-RADS 4 and 5 breast masses. Diagn Interv Imaging. 2016;97(9):883–9.

35.

Zheng

, Huang

, Liu

, et al. Shear-Wave Elastography of the Breast: Added Value of a Quality Map in Diagnosis and Prediction of the Biological Characteristics of Breast Cancer. Korean J Radiol. 2020;21(2):172–80.

36.

Tekcan Sanli

, Yildirim

, Kandemirli

, Sanli

, Aribal

. Evaluation of Multiparametric Shear Wave Elastography Indices in Malignant and Benign Breast Lesions. Acad Radiol. 2022;29(Suppl 1):S50–61.

37.

Zhou

, Zhan

, Chang

, et al. Breast lesions: evaluation with shear wave elastography, with special emphasis on the “stiff rim” sign. Radiology. 2014;272(1):63–72.

38.

Yang

, Xu

, Zhao

, Yin

, Chen

, Huang

. The role of tissue elasticity in the differential diagnosis of benign and malignant breast lesions using shear wave elastography. BMC Cancer. 2020;20(1):930.

39.

Chou

, Baikpour

, Zhang

, Mercaldo

, Lehman

, Samir

. Shear-Wave Elastography of the Breast: Impact of Technical Image Quality Parameters on Diagnostic Accuracy. AJR Am J Roentgenol. 2021;216(5):1205–15.

40.

Bartolotta

, Orlando

AAM

, Dimarco

, et al. Diagnostic performance of 2D-shear wave elastography in the diagnosis of breast cancer: a clinical appraisal of cutoff values. Radiol Med . 2022;127(11):1209–20.

41.

Zardavas

, Irrthum

, Swanton

, Piccart

. Clinical management of breast cancer heterogeneity. Nat Rev Clin Oncol. 2015;12(7):381–94.

42.

Samani

, Zubovits

, Plewes

. Elastic moduli of normal and pathological human breast tissues: an inversion-technique-based investigation of 169 samples. Phys Med Biol. 2007;52(6):1565–76.

43.

Park

, Kwon

, Woo

, et al. A Prospective Study on the Value of Ultrasound Microflow Assessment to Distinguish Malignant from Benign Solid Breast Masses: Association between Ultrasound Parameters and Histologic Microvessel Densities. Korean J Radiol. 2019;20(5):759–72.

44.

Kapetas

, Clauser

, Woitek

, et al. Quantitative Multiparametric Breast Ultrasound: Application of Contrast-Enhanced Ultrasound and Elastography Leads to an Improved Differentiation of Benign and Malignant Lesions. Invest Radiol. 2019;54(5):257–64.

45.

Lichtenbeld

, Barendsz-Janson

, van

Essen H

, Struijker Boudier

, Griffioen

, Hillen

. Angiogenic potential of malignant and non-malignant human breast tissues in an in vivo angiogenesis model. Int J Cancer. 1998;77(3):455–9.

46.

Hashizume

, Baluk

, Morikawa

, et al. Openings between defective endothelial cells explain tumor vessel leakiness. Am J Pathol. 2000;156(4)1363–80. doi: 10.1016/S0002-9440(10)65006-7.

47.

, Xu

, Guo

, et al. Combination of two-dimensional shear wave elastography with ultrasound breast imaging reporting and data system in the diagnosis of breast lesions: a new method to increase the diagnostic performance. Eur Radiol. 2016;26(9):3290–300.

48.

Thomas Stavros

, Classic Diagnosis of Breast Ultrasound (Chinese Translation).Wang ZL,translated. Beijing: Science Press, 2017.

49.

Weind

, Maier

, Rutt

, Moussa

. Invasive carcinomas and fibroadenomas of the breast: comparison of microvessel distributions–implications for imaging modalities. Radiology. 1998;208(2):477–83.