Advancing Early Breast Cancer Diagnosis and Ki‐67 Expression Prediction With Ultrasound Viscoelastography: Based on Peritumoral Assessment Authors: Xiaoli Huang, Bing Zhang, Fei Chen, Chunxiang Wei, Bei Lan, Huihui Yang, Demao Deng, Quanqing Zou Published: 2026 ABSTRACT ABSTRACTBackgroundBreast cancer is the most common invasive malignancy in women. Ultrasound viscoelastography may improve non‐invasive diagnosis, but the clinical relevance of peritumoral viscoelastic parameters (VPs) and their correlation with Ki‐67 remain unclear. This study explores these associations.MethodsA prospective study enrolled 256 breast masses (156 malignant, 100 benign). Patients underwent conventional ultrasound and viscoelastography. VPs were acquired within masses (A), peritumoral shells (Shell1: 1 mm, Shell2: 2 mm, Shell3: 3 mm), and combined regions (A' = A + Shell). Quantitative metrics included elasticity (V/1 E, V/0 E), dispersion coefficient (V/1 D), viscosity coefficient (V/0 V), and their maximum (Max), minimum (Min), mean(Mean), and standard deviation (SD) values. Diagnostic performance was evaluated by analyzing the receiver operating characteristic (ROC) curve and calculating the area under the curve (AUC), followed by logistic regression. VPs were compared between high/low Ki‐67 groups.ResultsThe AUCs of peritumoral VPs exceeded those from masses. V/1 D showed higher AUC than V/0 V and elasticity values, with Shell3 V/1 D SD having the highest AUC (0.840). The diagnostic efficacy of the peritumoral viscoelastic model surpassed that of the single elasticity parameter model, with 89.1% sensitivity, 73% specificity, 82.8% accuracy. Key variables included Shell3 V/1 D and Shell3 V/0 E. Shell1 V/0 V Max was significantly higher in high versus low Ki‐67 expression groups (p < 0.05).ConclusionThe viscoelastic model improves diagnostic accuracy by quantifying peritumoral biomechanics, and certain VPs may predict Ki‐67 expression. Peritumoral VPs show potential as novel non‐invasive imaging biomarkers for early breast cancer evaluation. Abbreviations Accuracy Area Under the Curve Dispersion Coefficient Model Elasticity Model Maximum Value Mean Value Minimum Value Main Variables Receiver Operating Characteristic Region of Interest Standard Deviation Strain Elastography Strain Elastography Specificity Shear Wave Elastography Viscoelastic Parameters Voigt Model Prefix Elasticity Value (based on the simple Voigt model) Viscosity Coefficient Linear Model Prefix Dispersion Coefficient Elasticity Value (based on linear model fitting) Viscosity Coefficient Model Introduction Breast cancer is the most common invasive cancer in women worldwide and the leading cause of cancer deaths among them [1]. Therefore, early diagnosis and intervention are of critical clinical importance. Current imaging methods for breast cancer include conventional ultrasound, ultrasound elastography (shear wave elastography [SWE] and strain elastography [SE]), contrast‐enhanced ultrasound, mammography, and magnetic resonance imaging. Among these, ultrasound elastography has been recognized by multiple clinical studies for its value in diagnosing breast masses due to its ability to reflect the mechanical properties of tissues [2, 3, 4, 5]. Its principle lies in that malignant lesions often exhibit higher stiffness due to increased cell density and connective‐tissue proliferation, which leads to tissue hardening [6]. However, traditional SWE simplifies breast tissue as a purely elastic material, neglecting its significant viscosity and dispersion, which may result in biased elasticity assessments [7]. Compared with SWE, viscoelastography introduces viscoelastic parameters (VPs) to model breast tissue, as a viscoelastic material and better reflects its intrinsic tissue properties. Previous studies have shown that VPs can differentiate between benign and malignant breast tumors [7, 8, 9, 10]. Cloutier G et al [7]. found in a study of 28 breast masses that the Voigt model shear elasticity, shear viscosity, and time constant of malignant masses were higher than those of the benign group, preliminarily confirming the diagnostic potential of viscosity‐related parameters. However, these studies were mostly exploratory with small sample sizes, and most of the VPs reported in studies [7, 8, 9] were obtained through post‐processing. A recent prospective multicenter study [11] compared the diagnostic efficacy of two VPs for breast cancer: the dispersion slope and the viscosity coefficient in the Voigt model, showing that the latter had a relative advantage. The study also confirmed that combining quantitative VPs with conventional ultrasound BI‐RADS classification can significantly enhance the ability of ultrasound to diagnose breast cancer. Overall, research on evaluating breast cancer with viscoelastography remains limited, particularly in systematically comparing real‐time parameters and developing optimized models. Thus, it is crucial to further expand the sample size and include diverse populations to clarify the diagnostic value of viscoelastography and promote its clinical application. Traditionally, it is believed that malignant breast masses are stiffer than benign ones. However, when necrosis occurs within breast cancer or in special types of breast cancer (such as mucinous carcinoma), the interior may appear softer. Therefore, it is inaccurate to evaluate only the internal elasticity of breast masses. Recent studies have found that the peritumoral tissue of malignant breast masses exhibits stiffness similar to that of the interior of the masses [12], and this stiffness is higher than that of benign breast lesions [13, 14]. This is a manifestation of cancer cell infiltration into the peritumoral tissue. In addition, tumor infiltration around the tumor in breast cancer patients without lymph node metastasis is an independent prognostic factor, significantly associated with an increased risk of tumor recurrence and mortality [12]. Therefore, objective measurement of peritumoral viscoelasticity in breast masses may be of greater clinical importance. Breast cancer occurs due to the uncontrolled proliferation of breast epithelial cells caused by carcinogenic factors [1]. Ki‐67 is a nuclear antigen that reflects tumor proliferation activity and regulates the cell proliferation process; It is widely recognized as the basis of cell reproduction [15]. Studies [16] have shown that the positive expression of Ki‐67 antigen is an important indicator of breast cancer biological behavior and is closely associated with lymph node metastasis and histological grading. This provides significant guidance for formulating post‐operative treatment plans for breast cancer. Although elastography studies have revealed correlations between elasticity scores, elasticity values, and Ki‐67 expression levels and proliferation indices [17, 18, 19], the relationship between VPs and Ki‐67 expression levels remains unclear, lacking relevant research reports. Therefore, this study utilized ultrasound viscoelastography to assess the viscoelastic properties of breast masses and peritumoral tissue. The technique enables real‐time quantification of the shear elastic modulus, dispersion coefficient (V/1 D, derived from linear model fitting), and viscosity coefficient (V/0 V, calculated using the simplified Voigt model). Optimal VPs were determined and integrated into a combination model. The diagnostic efficacy of this combination model was then compared with models based on single elasticity parameters. Additionally, correlations between VPs and Ki‐67 expression levels in breast cancer were analyzed. The study aims to evaluate the diagnostic value of peritumoral VPs in breast cancer and explore their relationship with tumor biological characteristics. These findings may provide evidence for optimizing non‐invasive preoperative diagnosis and postoperative assessment strategies for breast cancer. Materials and Methods Patients This study was approved by the ethics committee of the People's Hospital of Guangxi Zhuang Autonomous Region (approval number: KY‐KJT‐2023‐148), and informed consent was obtained from each patient. A total of 230 female patients with 262 breast masses were enrolled. These patients were treated at the Breast Surgery Department of the People's Hospital of Guangxi Zhuang Autonomous Region from August 2023 to February 2025. The inclusion criteria were as follows: patients aged 18 years or older; patients who had undergone conventional ultrasound and viscoelastography examinations before surgery; patients who underwent biopsy or surgical resection within 2 weeks after the ultrasound examinations and obtained postoperative histopathological results; and breast masses that were solid or predominantly solid as detected by ultrasound. The exclusion criteria included cystic or predominantly cystic solid‐cystic mixed masses; masses with extensive coarse calcification; masses with a maximum diameter greater than 5.0 cm; patients with breast implants; patients who had undergone ipsilateral breast surgery, neoadjuvant chemotherapy, or systemic chemoradiotherapy; and those who did not obtain satisfactory viscoelastic images. This study ultimately included 226 patients, aged 18 to 84 years (mean age 49.87 ± 13.09 years), with a total of 256 masses. Among them, two patients with a total of two masses were excluded due to poor image quality. Additionally, two patients with a total of four masses were excluded for lack of histopathological results. Image Acquisition The breast ultrasound examination was performed using a Mindray Resona 9 ultrasound diagnostic device, with a probe model of L11‐3W. After routinely scanning the patient's breast and selecting the target breast nodule, the viscoelastic mode was activated. The region of interest (ROI) was chosen to maximize the mass's cross‐sectional area, and the mass was placed in the center of the sampling frame. If the mass was large, the cross‐sectional area was appropriately adjusted to ensure that the mass edge was at least 3 mm away from the sampling frame. Once the frame was set, the patient was instructed to hold her breath. When the M‐STB Index—a rating scale where higher stars indicate better image quality—was above four stars and the quality map was uniformly green with no purple artifacts in the background, the image was considered qualified and saved. All examinations were performed by an experienced radiologist with over 10 years of experience in breast ultrasound and elastography. The radiologist had received training in the implementation of viscoelastography technology and standardized operational procedures. Image Analysis The viscoelastic image displayed four types of maps (Figure 1A): a grayscale image, an elastogram, a quality control map, and a viscoelasticity map. In the elastogram, red represented hard tissue, while blue indicated soft tissue. In the viscoelasticity map, the orange–white gradient corresponded to highly viscous tissue, while the dark red–black gradient indicated less viscous tissue. The mass was manually outlined on the grayscale image, defining the intratumoral ROI denoted as A. First, the dispersion coefficient measurement mode (based on linear model fitting, denoted by the prefix V/1) was selected. The elasticity values (V/1 E: V/1 E Max, Mean, Min, SD) and the dispersion coefficient values (V/1 D: V/1 D Max, Mean, Min, SD) within A were recorded (Figure 1A–C). The mode was then switched to acquire parameters based on the simple Voigt model (denoted by the prefix V/0). The elasticity values (V/0 E: V/0 E Max, Mean, Min, SD) and the viscosity coefficient values (V/0 V: V/0 V Max, Mean, Min, SD) within A were subsequently recorded (Figure 1D‐F). At this point, the measurement of VPs for A were complete, totaling 16 parameters. Subsequently, the ultrasound system's “Shell” function was utilized to automatically generate three concentric peritumoral shells at distances of 1 mm, 2 mm, and 3 mm from the border of A, which were denoted as Shell1, Shell2, and Shell3, respectively. Combined regions were defined as A' (where A' = A + Shell, and Shell referred to any one of Shell1, Shell2, or Shell3). The same set of VPs were sequentially measured within each peritumoral shell and each combined region following the protocol described above, resulting in a total of 112 parameters across all regions (Figure 1). Viscoelastic image. (a) Grayscale image; (b) Elastogram; (c) Quality control map; (d) Viscoelasticity map. (A–F) Red boxes indicate Shell ranges. (A–C) V/1 D and V/1 E measurement charts. (C) Yellow box shows V/1 E values (Mean, Max, Min, SD) for A, Shell1, and A' (A+Shell1); Blue box shows V/1 D values (Mean, Max, Min, SD) for A, Shell1, and A' (A+Shell1). (D–F) V/0 V and V/0 E measurement charts. (D) Yellow box shows V/0 E values (Mean, Max, Min, SD) for A, Shell3, and A' (A+Shell3); Blue box shows V/0 V values (Mean, Max, Min, SD) for A, Shell3, and A' (A+Shell3). Abbreviation: V(m/s/kHz), V/1 D; V(Pa·s), V/0 V. All images and data analyses were saved on the ultrasound device's hard drive and then independently analyzed by the examining physician after the examination. The Reproducibility of VPs The radiologist conducted intra‐observer reproducibility assessment on 30 breast masses from 30 patients. After completing viscoelastography training, the radiologist performed two repeated viscoelastic imaging sessions on the same patient's breast mass, with an interval exceeding 10 min. All VPs were then recorded for subsequent calculation of intra‐observer reproducibility. Pathological Examination The final pathological diagnosis was based on the histopathology of tissues obtained from surgical resection or biopsy. The breast pathology specialist was blinded to the ultrasound results when making the diagnosis. Ki‐67 positive cells were defined as those with pale yellow or brownish‐yellow nuclear staining. In this study, the average proportion of positive cells was calculated by analyzing ten randomly selected fields of view at 400 × magnification. Based on the expert consensus from the 2011 St. Gallen International Breast Cancer Conference [20], a cutoff value of 14% was applied. Low expression was defined as < 14% positivity, and high expression was defined as ≥ 14% positivity. Statistical Analysis Statistical analysis was performed using SPSS 27.0. Normally distributed data were presented as mean ± standard deviation and analyzed using the independent sample t‐test. Non‐normally distributed data were expressed as median with interquartile range and analyzed using the Mann‐Whitney U test. Enumeration data were expressed as number of cases or percentages. Postoperative pathology results were used as the gold standard, with the pathological outcome of masses (benign/malignant) as binary classification labels. The receiver operating characteristic (ROC) curves for VPs were plotted, and the area under the curve (AUC) was calculated. Diagnostic value was classified based on AUC as high (AUC > 0.9), moderate (0.7 < AUC ≤ 0.9), low (0.5 ≤ AUC ≤ 0.7), or no diagnostic value (AUC < 0.5). Selected parameters from ROC curves were analyzed individually or jointly using binary logistic regression to determine the sensitivity (Se), specificity (Sp), and accuracy (Ac) of the models. A p‐value < 0.05 was considered statistically significant. The intra‐class correlation coefficient (ICC) was used to evaluate the reproducibility of radiologists' measurements. Results Pathological Diagnoses This study included a total of 256 breast masses, comprising 156 malignant and 100 benign masses. Among the malignant masses, non‐specific invasive carcinoma was the most common (118 cases, 75.6%). Among the benign masses, fibroadenoma was predominant (39 cases, 39.0%) (see Table 1). Table 1Pathological diagnoses in 256 breast masses.Malignant N Benign N Invasive Carcinoma of No Special Type (NST)118Fibroadenoma39Invasive Lobular Carcinoma10Fibrocystic Changes18Ductal Carcinoma In Situ (DCIS)6Mammary Hyperplasia13Mucinous Carcinoma5Adenosis of Breast9Microinvasive Carcinoma4Intraductal Papilloma8Paget's Disease1Mastitis4Neuroendocrine Tumor1Fibroepithelial Tumor of Breast3Breast Carcinoma with Apocrine Differentiation1Diabetic Mastopathy2Adenoid Cystic Carcinoma1Duct Ectasia3Metastatic Carcinoma1Lipoma1Mixed Carcinoma8John Wiley & Sons, Ltd. Pathological diagnoses in 256 breast masses. ROC Comparison of VPs 1. Comparison of breast mass and peritumoral tissue parameters: The AUCs of most peritumoral shells were greater than those of A and A'. Shells1, 2, and 3 each had moderate diagnostic value in distinguishing breast cancer, with Shell3 exhibiting the highest AUC. 2. Comparison of elasticity, dispersion coefficient and viscosity coefficient values: V/1 E, V/0 E, V/0 V, and V/1 D all demonstrated diagnostic value, with V/1 D showing the largest AUC. Among all V/1 D values, Shell3 V/1 D SD achieved the largest AUC of 0.840. The Shell3 V/1 D SD values were (3.49 ± 0.83) m/s/kHz in the malignant group and (2.34 ± 0.78) m/s/kHz in the benign group. Regarding elasticity values, Shell3 V/0 E Max had the greatest AUC (0.813). Its values were 107.13 (75.08–161) KPa in the malignant group and 52.74 (35.75–78.77) KPa in the benign group. For V/0 V values, Shell3 V/0 V Max demonstrated the highest AUC (0.813), with values of (12.59 ± 4.60) Pa·s in the malignant group and (7.20 ± 3.64) Pa·s in the benign group (Figure 2). The ROC results for the main VPs are presented in Table 2. VPs (except for the Min value) were higher in breast cancer patients than in benign mass patients. (A) V/1 D and V/1 E measurements in invasive carcinoma of no special type (NST). (B) V/0 V and V/0 E measurements in NST. (C) V/1 D and V/1 E measurements in intraductal papilloma. (D) V/0 V and V/0 E measurements in intraductal papilloma. Table 2Main parameters ROC results.ParametersAUCA V/1 E Max0.694Shell1 V/1 E Max0.745A' (A+Shell1) V/1 E Max0.732A V/0 E Max0.705Shell1 V/0 E Max0.770A' (A+Shell1) V/0 E Max0.743Shell2 V/1 E Max0.791A' (A+Shell2) V/1 E Max0.778Shell2 V/0 E Max0.795A' (A+Shell2) V/0 E Max0.786Shell3 V/1 E Max0.801A' (A+Shell3) V/1 E Max0.785A' (A+Shell3) V/0 E Max0.794Shell3 V/0 E Max0.813Shell3 V/0 E Min0.417Shell3 V/0 E Mean0.744Shell3 V/0 E SD0.806A V/1 D Max0.734A V/0 V Max0.706A' (A+Shell1) V/1 D Max0.787Shell1 V/1 D Max0.802Shell1 V/1 D SD0.780A' (A+Shell1) V/0 V Max0.754Shell1 V/0 V Max0.784Shell1 V/0 V SD0.760A' (A+Shell2) V/1 D Max0.801Shell2 V/1 D Max0.819Shell2 V/1 D SD0.814A' (A+Shell2) V/0 V Max0.791Shell2 V/0 V Max0.801Shell2 V/0 V SD0.792A' (A+Shell3) V/1 D Max0.809Shell3 V/1 D Max0.825Shell3 V/1 D Min0.500Shell3 V/1 D Mean0.676Shell3 V/1 D SD0.840A' (A+Shell3) V/0 V Max0.799Shell3 V/0 V Max0.813Shell3 V/0 V Min0.464Shell3 V/0 V Mean0.741Shell3 V/0 V SD0.804John Wiley & Sons, Ltd. Main parameters ROC results. Variables and Diagnostic Efficacy of the Breast Cancer Diagnosis Model Based on the ROC analysis, specific variables were selected for inclusion in the binary logistic regression equation. The parameters of Shell3 V/1 D and Shell3 V/0 E were used as the main variables (MV) to develop the optimal viscoelastic model, termed the peritumoral viscoelastic model (Se = 89.1%, Sp = 73%, Ac = 82.8%). Moreover, the diagnostic performance and MV of other models were as follows: for the elasticity model (E model), Se = 82.1%, Sp = 64.0%, Ac = 75.0%, the MV was Shell3 V/0 E; for the dispersion coefficient model (D model), Se = 83.3%, Sp = 72.0%, Ac = 78.9%, the MV was Shell3 V/1 D; and for the viscosity coefficient model (V model), Se = 80.8%, Sp = 61.0%, Ac = 73.0%, the MV was Shell3 V/0 V. The detailed variables for each diagnostic model are summarized in Table 3. Table 3Variables in the Equation.BStd. ErrorWalddf p valueVE modelShell3 V/1 D Mean−0.6280.2059.37710.002Shell3 V/1 D SD1.9620.37826.9131< 0.001Shell3 V/0 E Max0.0100.0047.42910.006Shell3 V/0 E Min−0.1570.714.88710.027Constant−3.9660.69332.7951< 0.001E modelShell3 V/0 E Max0.0090.0062.03610.154Shell3 V/0 E Min−0.2080.0699.23410.002Shell3 V/0 E SD0.0980.0464.58810.032Constant−1.9520.47616.7951< 0.001D modelShell3 V/1 D Mean−0.6820.19212.6361< 0.001Shell3 V/1 D SD2.4450.33652.7991< 0.001Constant−4.3690.64246.3331< 0.001V modelShell3 V/0 V Max0.3050.04251.9541< 0.001Constant−2.4840.40637.5011< 0.001 Note: B, Regression coefficient; df, degree of freedom.John Wiley & Sons, Ltd. Variables in the Equation. Note: B, Regression coefficient; df, degree of freedom. Differences in VPs between High and Low Ki‐67 Expression Groups in Breast Cancer Compared to the Ki‐67 low‐expression group, the maximum value of A, Shells, and A' were elevated in the high‐expression group (p < 0.05). Among these parameters, the inter‐group difference in Shell1 V/0 V Max was the most pronounced, with the value in the Ki‐67 high‐expression group (11.85 ± 4.53 Pa·s) significantly higher than that in the low‐expression group (8.36 ± 4.04 Pa·s) (Cohen's d = 0.787) (Figure 3). A representative case of breast cancer demonstrating elevated Shell1 V/0 V Max values associated with high Ki‐67 expression. (A) NST in a patient with Ki‐67 ≥ 14%, showing a Shell1 V/0 V Max value of 19.07 Pa·s (indicated by red box). (B) NST in a patient with Ki‐67 < 14%, showing a Shell1 V/0 V Max value of 7.74 Pa·s (indicated by red box). The Reproducibility of VPs The optimal VPs, Shell3 V/1 D SD, showed excellent reproducibility between two measurements performed by radiologist, with an ICC of 0.823. Discussion Salient observations from this study include: (1) The peritumoral VPs exhibited higher diagnostic potential for breast cancer compared to those of the masses themselves. Among these parameters, V/1 D performed better than V/0 V, V/1 E and V/0 E. The exploratory peritumoral viscoelastic model using Shell3 V/1 D and Shell3 V/0 E as variables, showed promising diagnostic performance within our cohort. (2) Furthermore, the VPs values in the high Ki‐67 expression group were higher than those in the low expression group, with the most significant inter‐group difference observed in Shell1 V/0 V Max, suggesting that this parameter may predict Ki‐67 expression levels. Collectively, these findings indicate that focusing on the peritumoral VPs and constructing a combined model not only improves the detection efficiency of breast cancer but also provides a new exploratory direction for reducing unnecessary biopsies in clinical practice. Based on these ultrasound viscoelastography indicators, we propose a new approach that merits further investigation for both early diagnosis of breast cancer and noninvasive assessment of Ki‐67 status. In this study, by comparing the AUC of the ROC curves for VPs, we found that the peritumoral VPs are more valuable in diagnosing the benignity or malignancy of breast masses than those of the masses themselves, especially for the tissue within a 3 mm range around the mass. Furthermore, the VPs (except for the Min value) were consistently higher in the malignant group than in the benign group (Figure 2). This result aligns with findings from previous studies. For instance, Jia W et al. [11] noted that the diagnostic efficacy of peritumoral VPs is superior to that of the masses themselves. Park HS et al. [21] measured the elasticity value of the internal, peripheral, and surrounding stromal tissues of masses using SWE and found that the maximum elasticity values occurred in the surrounding stromal tissue approximately 2 to 4 mm from the tumor edge. This phenomenon may be attributed to the infiltration of breast cancer cells, which leads to the formation of abnormally stiff collagen fibers associated with cancer fibroblasts in the peritumoral tissue, thereby increasing the stiffness of the tissue surrounding malignant masses [22]. Drawing on Sugimoto K et al.'s [23] findings regarding multiparametric ultrasound in the liver, we speculate that after malignant cell infiltration into breast tissue, tumor inflammatory cells appear in and around the mass, causing changes in tissue viscosity. Moreover, the inflammatory response induced by malignant infiltration is more uniformly significant in the peritumoral tissue, resulting in highly consistent changes in the VPs of that area. Therefore, quantitative measurement of peritumoral VPs is more diagnostically valuable than intratumoral parameters. Additionally, the AUC of V/1 D SD was higher than that of V/1 D Max in Shell3, which remained optimal in Shell1 and Shell2. This may be because SD reflects the dispersion of data, meaning that the SD of breast masses may represent the differential distribution of tissue cells. Shell3 covers a larger range and is farther from the mass center, potentially encompassing a mixed distribution of normal and malignant breast tissues. Due to the heterogeneity of breast malignant tissues, the differences between tissue cells are more pronounced than in Shell1 and Shell2, leading to earlier indication of abnormalities in the SD value compared to the Max value in Shell3. In viscoelastic breast tissue, the shear wave velocity varies with frequency, leading to differences in elasticity values measured at different frequencies. This study innovatively compared the elasticity values derived from the simple Voigt model (V/0 E) with those obtained from linear model fitting (V/1 E) for the differential diagnosis of breast cancer. The results showed that the AUC of V/0 E was higher than that of V/1 E. This difference may stem from variations in model algorithms and frequency responses. The viscosity parameters (V/1 D and V/0 V) serve as indicators of viscosity, which relates to tissue composition and reflects the magnitude of internal friction generated by inter‐particle interactions. Their measurements are not affected by changes in frequency, which theoretically allows for a more accurate assessment of breast masses characteristics. In this study, all VPs—V/1 E, V/0 E, V/1 D, and V/0 V—showed diagnostic value. Among them, V/1 D exhibited the highest AUC, which is consistent with theoretical expectations. V/1 D can be analyzed as a slope [9, 10] and estimated using a linear model fit to describe the variation of shear wave velocity with frequency. V/0 V, on the other hand, is based on the simple Voigt model (whose equation includes factors such as viscosity, medium density, and elasticity) to illustrate frequency‐dependent velocity. Previously, V/1 D was primarily applied in studies of liver lesions [23, 24, 25], but had limited application in breast diagnosis; V/0 V was even more rarely used clinically. Jia W et al. [11] first compared the efficacy of these two parameters in breast cancer differentiation and concluded that V/0 V has higher diagnostic value than V/1 D. However, our findings contradicted this conclusion, potentially due to two factors. First, although the Voigt model employs a more complex equation than the linear model, it still imperfectly represents biological tissue; thus, the diagnostic superiority of either parameter requires further investigation. Second, discrepancies in study populations may contribute to inconsistent results, necessitating validation with larger, more diverse cohorts. Previous studies have suggested that SWE parameters can complement conventional ultrasound to improve differentiation between benign and malignant breast masses [26, 27]. In this study, by screening VPs, we constructed a viscoelastic combined model for diagnosing breast cancer, as well as single elasticity parameter models (E, D, and V models). The results showed that the viscoelastic model based on the value of V/1 D and V/0 E within the 3 mm peritumoral tissue exhibited significantly superior diagnostic performance compared to the single elasticity parameter models. As shown in Table 3, the key elasticity variables in the optimal viscoelastic model were Shell3 V/0 E Max and Shell3 V/0 E Min. This result is similar to that of Yang H et al. [13], who found that Emax and Emin of the 3 mm tissue surrounding the mass are significant predictors of malignancy. However, the mean values of elasticity (V/1 E and V/0 E) in this study did not demonstrate optimal diagnostic value, which differs from previous studies [28, 29] that suggested Emax and Emean as the optimal discriminant parameters. This discrepancy may be attributed to differences in ROI size and statistical methods [30]. Finally, multivariate analysis indicated that the combined application of Shell3 V/1 D (Mean, SD) and Shell3 V/0 E (Max, Min) was most diagnostically valuable in differentiating between benign and malignant breast masses using viscoelastography. Ki‐67, a nuclear protein specifically expressed in proliferating cells, is a key biomarker for assessing the proliferative activity of breast cancer. Its high expression is associated with stronger aggressiveness and poor prognosis [31]. This study is the first to explore the correlation between VPs and Ki‐67 expression, finding that the maximum values of VPs (V/1 E, V/0 E, V/1 D, V/0 V) in masses, tissues within 1 mm, 2 mm, and 3 mm around the masses and combined regions were significantly higher in the high Ki‐67 expression group than in the low Ki‐67 expression group (p < 0.05). Among these, the maximum value of V/0 V derived from the 1 mm peritumoral tissues demonstrated a particularly significant intergroup difference (Cohen's d = 0.787, close to Cohen's recommended criterion for a large effect size threshold of 0.8), suggesting that this noninvasive parameter has potential value in predicting Ki‐67 expression. This observation is likely due to pathological changes in the peritumoral microenvironment: high Ki‐67 expression promotes active degradation and remodeling of the extracellular matrix, accompanied by significant inflammatory infiltration, abnormal angiogenesis, high‐permeability exudation, and interstitial edema [32, 33]. These factors collectively increase the viscous resistance of the extracellular space, leading to an increase in the V/0 V of peritumoral tissues detected by ultrasound. Therefore, the V/0 V of peritumoral tissues (especially Shell1 V/0 V Max) may be regarded as a quantifiable “imprint” of high tumor proliferative activity on the physical properties of the local microenvironment, providing a new imaging biomarker approach for noninvasive assessment of tumor proliferative status. However, its clinical application value needs further validation in prospective large‐sample studies. This preliminary viscoelastography application study has several limitations. First, we focused solely on quantitative VPs and did not perform qualitative diagnostic analysis or comparative assessments with conventional ultrasound or traditional elastography. Second, the relatively small sample size precluded subgroup analyses (e.g., based on histological subtypes, grading, or lesion size, age, menstrual status, and breast density), potentially impacting the viscoelastography findings. Additionally, conducting viscoelastography examinations subsequent to conventional ultrasound assessments may have introduced observer bias in elastographic image acquisition and interpretation by radiologists. Conclusion This study developed a diagnostic model that integrated peritumoral VPs—specifically the dispersion coefficient and elasticity value based on the simple Voigt model—significantly enhancing the accuracy of breast cancer diagnosis. Furthermore, the peritumoral viscosity coefficient emerged as a novel non‐invasive imaging biomarker, demonstrating initial potential for predicting tumor biological behavior. These findings indicate the considerable clinical promise of peritumoral VPs for the early diagnosis of breast cancer and the development of personalized postoperative treatment strategies. Author Contributions Xiaoli Huang: conceptualization, methodology, writing – original draft. Bing Zhang: methodology, formal analysis. Fei Chen: methodology, formal analysis. Chunxiang Wei: methodology, formal analysis. Bei Lan: methodology, data curation. Huihui Yang: formal analysis. Demao Deng: writing – review and editing. Quanqing Zou: writing – review and editing. Ethics Statement This study was conducted in accordance with the ethical principles of the Declaration of Helsinki. Ethical approval was obtained from the Ethics Committee of the People's Hospital of Guangxi Zhuang Autonomous Region (protocol number: KY‐KJT‐2023‐148). Written informed consent was obtained from all individual participants included in the study. Conflicts of Interest The authors declare no conflicts of interest. Transparency Statement The Corresponding author (Demao Deng) affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.