Utility of ultrasound Angio-PLUS imaging for detecting blood flow in breast masses and comparison with color Doppler for differentiating benign from malignant masses

Abstract

Background

Tumor neo-angiogenesis plays an important role in the development and growth of breast cancers, but its detection by imaging is challenging. A novel microvascular imaging (MVI) technique, Angio-PLUS, promises to overcome the limitations of color Doppler (CD) in detecting low-velocity flow and small diameter vessels.

Purpose

To determine the utility of the Angio-PLUS technique for detecting blood flow in breast masses and compare it with CD for differentiating benign from malignant masses.

Material and Methods

A total of 79 consecutive women with breast masses were prospectively evaluated using CD and Angio-PLUS techniques, and biopsied as per BI-RADS recommendations. Vascular imaging scores were assigned using three factors (number, morphology, and distribution) and vascular patterns were divided into five groups: internal-dot-spot, external-dot-spot, marginal, radial, and mesh patterns. The independent samples t-test, Mann–Whitney U test, Wilcoxon signed rank test, or Fisher’s exact test were used to compare the two groups as appropriate. Area under the receiver operating characteristic (ROC) curve (AUC) methods were used to assess diagnostic accuracy.

Results

Vascular scores were significantly higher on Angio-PLUS than CD (median=11, [IQR=9–13] vs. 5 [IQR=3–9], P < 0.001). Malignant masses had higher vascular scores than benign masses on Angio-PLUS (P < 0.001). AUC was 80% (95% CI=70.3–89.7; P < 0.001) for Angio-PLUS and 51.9% for CD. Using Angio-PLUS at a cutoff value of ≥9.5, sensitivity was 80% and specificity was 66.7%. Vascular pattern descriptors on AP showed good correlation with histopathological results (PPV mesh 95.5%, radial 96.9%, and NPV of marginal orientation 90.5%).

Conclusion

Angio-PLUS was more sensitive in detecting vascularity and superior in differentiating benign from malignant masses compared to CD. Vascular pattern descriptors on Angio-PLUS were useful.

Keywords

Microvascular imaging color Doppler Angio-PLUS breast cancer malignant breast mass benign breast mass

Introduction

Globally, breast cancer is the leading cancer among women with an estimated 2.3 million new cases reported every year, representing 11.7% of all cancer cases worldwide (1). Tumor neo-angiogenesis, defined as the formation of new vessels in pre-existing vascular networks, plays an important role in the development, growth, and metastasis of breast cancers (2,3). The high density of these abnormal micro-vessels is associated with an increased likelihood of malignancy and can be measured via histopathological examination of tissue specimens. Microscopic findings include irregular and variable vessel caliber, reticular and disordered vascular networks with vessel tortuosity, arteriovenous shunts, and incomplete vascular walls (4).

Indirect assessment of tumor vascularity by non-invasive imaging techniques such as color Doppler (CD) has been attempted but was not found to be reliable (5 –9) as it provides information only about macro-vessels, with low sensitivity for detection of angiogenic micro-vessels (diameter <0.1 mm) (10,11). CD findings do not sufficiently correlate with micro-vessel density on pathological examination and have limited utility in differentiating small or hypo-vascular cancers from hyper-vascular benign lesions (11,12).

Ultrasonic Doppler signals are produced by flowing blood, as well as tissue motion in nearby structures (e.g. patient movement, pulsation, or respiratory movements). These unwanted “clutter” signals overlap the low-velocity blood flow signals. In conventional Doppler, a single-dimension wall filter is applied to remove clutter artifacts; however, this is unable to differentiate between motion artefacts and actual blood flow, hence it also removes the signals from low-velocity flow (13 –15). To overcome this problem, various ultrasound (US)-based microvascular imaging (MVI) techniques have been developed promising better visualization of low-velocity and small diameter blood vessel flow. One of the best known non-contrast microvascular US imaging techniques is Superb Microvascular Imaging (SMI) (15). Other innovative techniques available are Angio-PLUS (Supersonic Imagine, Alpes cote d' azur, France), MicroFlow imaging (Philips Healthcare, Pittsburg and Reedsville, North America), and MV-Flow (Samsung Medison, Seoul, Republic of Korea) (16). Some studies have investigated the utility of SMI in different breast lesions (17 –22).

In the Angio-PLUS (Plane Wave Ultra-Sensitive Imaging) technique, plane waves are sent into the body, and each pixel of explored tissue is interrogated continuously with a significantly higher sampling rate (5–10 times faster) than in color Doppler imaging (23).

We hypothesized that due to its higher sensitivity for detection of low velocity blood flow, the Angio-PLUS technique could improve assessment of vascularity of breast masses compared to color Doppler. Therefore, the aim of the present study was to determine the utility of ultrasound Angio-PLUS imaging for detecting blood flow in breast masses and to compare it with color Doppler for differentiating benign from malignant masses.

Material and Methods

The present study was a single-center, hospital-based, prospective, observational study conducted in the Department of Radio-diagnosis in collaboration with Departments of Endocrine and Breast Surgery and Pathology between November 2019 and June 2021. This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by the Ethics Committee of Sanjay Gandhi Post-Graduate Institute of Medical Sciences Lucknow (reference no. IEC 2020-61-MD-EXP-15, 16 April 2020).

Patients

Inclusion criteria were as follows: patients with breast masses (with the largest dimension ≤50 mm in size), who gave informed consent. Patients with a prior history of surgery or biopsy were excluded.

Sample size

A total of 79 women imaged for breast masses by US (including the color Doppler and Angio-PLUS techniques) and biopsied as per BI-RADS recommendations (24) were included.

The reference standard for diagnosis was histopathology.

Machine

All scans were performed using the Aixplorer® Ultrasound system (Supersonic Imagine, Aix-en-Provence, France) with a linear-array probe (SuperLinear™ SL10-2, bandwidth: 2–10 MHz) for the breast (using a preoptimized breast protocol). Sonographic measurements were performed by the same radiologist after being trained by an application specialist for performing Angio-PLUS and performing >50 scans under direct supervision. The stored images were jointly reviewed by two radiologists with 20 and 10 years of experience in US.

Scanning technique

Both breasts and axillae were scanned systematically (as per American Institute of Ultrasound in Medicine practice guidelines). The grayscale findings (location, size, shape, orientation, margins, echo-pattern, and posterior features) were documented for all masses along with color Doppler and Angio-PLUS findings. The parameters for color Doppler were as follows: velocity scale <2.5 cm/s; dynamic range = 20 dB; frame rate = 8–11 frames/s. The parameters for Angio-PLUS were as follows: velocity scale <2.5 cm/s; dynamic range = 21 dB; and frame rate = 27–60 frames/s. The plane with the richest vessels was selected as representative image for evaluation.

Image interpretation

Vascular imaging scores were assigned using a three-factor system to score the number, morphology, and distribution of tumor vessels (21). The number of vessels within a mass were counted and scored from 0 to 6 (≥6 vessels were assigned a score of 6).

Morphology was graded according to the complexity of the vessels, as follows: 1 = dot-like; 2 = linear; 3 = branching; and 4 = penetrating or shunting. A penetrating vessel was defined as a continuous signal extending from outside the lesion to inside the lesion. Shunting vessels were defined as capillary network connections from one or more different vessels with chaotic irregularity. The distribution of vessels was scored as follows: 1 = peripheral, defined as all vessels located at the margin or within 2 mm of the margin of the mass; 2 = central, defined as all vessels located within the lesion and not extending to the periphery; 3 = both. The overall vascular imaging score for each mass was recorded from 0 to 13. These scores were calculated on color Doppler and Angio-PLUS separately and then compared.

Vascular patterns were initially classified as described by Svensson et al. (25), comprising four major groups, and each group subcategorized according to vessel orientation (marginal, parallel, and radial).

Later, we modified the classification for simplicity and described five patterns (Figs. 1 and 2) as follows:

marginal-linear vessels parallel to margins of tumor (internal or external to the lesion);

radial-linear vessels in radial orientation (internal or external to the lesion);

internal dot-spot (IDS) – spot-like vessels within the mass;

external dot-spot (IDS) – spot-like vessels along the periphery of the mass;

mesh-like pattern of vessels with an internal branching pattern.

Fig. 1.

Images of the vascular patterns in three different masses: (a) marginal vessels; (b) external dot-spot (EDS) vessels in a fibroadenoma; and (c) internal dot-spot (IDS) vessels in a fibroadenoma.

Fig. 2.

Images of the vascular patterns in two different malignant masses: (a) radial vessels in a carcinoma; (b) disordered mesh-pattern of vessels in an invasive ductal carcinoma.

Statistical analysis

Variables were presented as mean ± standard deviation if data were normally distributed, otherwise median (interquartile range [IQR]) was used. For categorical variables, frequency and percentage were used. To compare the means and medians between the two groups, the independent samples t-test and Mann–Whitney U test were used, respectively. To compare the proportions between the two groups, Fisher’s exact test was used. Similarly, in the same patients (for all 79 patients), vascular scores between Angio-PLUS and CD were compared using the Wilcoxon signed rank test. The area under the receiver operating characteristics (ROC) curve (AUC) was used to find the diagnostic accuracy of Angio-plus and CD, and an appropriate cutoff value was identified. A cutoff value was considered only if both the sensitivity and specificity were >50%. For the chosen cutoff value, the corresponding sensitivity, specificity, positive likelihood ratio (LR+), and negative likelihood ratio (LR−) were calculated. P < 0.05 was considered to be statistically significant. SPSS version 23 (IBM Corp., Armonk, NY, USA) was used for the data analysis.

Results

A total of 79 consecutive women (mean age = 41.94 years; age range = 18–64 years) with breast masses (55 malignant, 24 benign) were enrolled in the study. Of the 24 benign lesions, 22 (91%) were fibroadenomas, 1 (4%) was fibrocystic disease, and 1 (4%) was granulomatous disease. Of the 55 malignant cases, 53 (96.4%) were infiltrating ductal carcinomas, and 1 (1.8%) case each of infiltrating lobular carcinoma and ductal carcinoma in situ. The mean age of patients with malignant masses was higher than those with benign masses (45.45 ± 10.18 years vs. 38.38 ± 9.94; P = 0.005).

Grayscale US analysis revealed that most of the benign lesions had an oval shape, circumscribed margins, iso to hypoechoic echotexture, and a parallel orientation. Most of the malignant lesions were irregular in shape, showed indistinct, micro-lobulated, angulated, and spiculated margins, and had an antiparallel orientation (Table 1).

Table 1.

Grayscale ultrasound features of lesions.

Descriptors	Histopathology
Descriptors	Total (n = 79)	Benign (n = 24)		Malignant (n = 55)	P value
Shape
Oval	21 (26.5)	18 (85.7)		3 (14.3)	<0.001
Round	4 (5.2)	2 (50)		2 (50)	0.581
Irregular	54 (68.3)	4 (7.4)		50 (92.6)	<0.001
Orientation
Parallel	43 (54.4)	24 (54.8)	19 (44.2)		<0.001
Antiparallel	36 (45.6)	0 (0)	36 (100)		<0.001
Margins
Circumscribed	21 (26.5)	18 (85.7)	3 (14.2)		<0.001
Indistinct	17 (21.5)	2 (11.7)	15 (88.2)		<0.001
Micro-lobulated	15 (18.9)	4 (23.5)	11 (73.3)		<0.001
Angulated	10 (12.6)	0 (0)	10 (100)		0.027
Spiculated	16 (20.4)	0 (0)	16 (100)		0.001
Echotexture
Hypo-echoic	30 (37.9)	15 (50)	15 (50)		0.019
Iso-echoic	4 (5.2)	3 (75)	1 (25)		0.081
Hetero-echoic	45 (59.9)	6 (13.3)	39 (86.7)		<0.001

Values are given as n (%). P values are compared using the chi-square test or Fisher’s exact test. P < 0.05 is significant.

Angio-Plus imaging was more sensitive in detecting vascularity in lesions (Figs. 3–5) and obtained significantly higher median vascular scores (for all 79 masses) compared to CD (median [IQR]: 11 [9–13] vs. 5 [3–9]; P < 0.001). On comparison of the median vascular imaging scores of benign versus malignant cases, it was observed that on Angio-PLUS imaging, there was a significantly higher vascular score in malignant masses compared to benign masses (P < 0.001), whereas on CD, there was no statistically significant difference between the vascular scores of the two groups (Table 2).

Fig. 3.

(a) CD showing external dot-spot (EDS) vessel in a fibroadenoma. (b) Angio-PLUS is more sensitive and detects a fine / slender radial vessel in the same mass, no tortuosity or color aliasing noted. Compare this with (c) tortuous radial vessels in a malignant mass. CD, color Doppler.

Fig. 4.

Invasive ductal carcinoma. (a) CD image on the left, showing few internal vessels in the mass; and (b) Angio-PLUS image of same mass (on the right), showing extensive disordered mesh of vessels within and around the mass, indicating the higher sensitivity of Angio-PLUS. CD, color Doppler.

Fig. 5.

Ductal carcinoma in situ, seen as an ill-defined area of hypo-echogenecity with internal echogenic specks denoting a non-mass finding on ultrasound (asymmetry with associated calcifications): (a) CD was unable to detect vascularity in the lesion; (b) Angio-PLUS is more sensitive and detected a disordered mesh of vessels within and around the lesion. CD, color Doppler.

Table 2.

Comparison of vascular scores between benign and malignant lesions by Angio-PLUS and color Doppler.

			Vascular score
Method	Histopathology	n	Vascular score		P value*
Angio-PLUS	Malignant	55	11.16 ± 1.84	11 (10–13)	<0.001
Angio-PLUS	Benign	24	8.54 ± 2.60	9 (8.3–10.8)	<0.001
Color Doppler	Malignant	55	5.65 ± 3.95	5 (3–9)	0.784
Color Doppler	Benign	24	5.25 ± 3.85	5.5(0.8–8.5)	0.784

Values are given as mean ± SD or median (IQR).

*Compared by Mann–Whitney U test. P < 0.05 is significant.

Diagnostic accuracy of Angio-PLUS for detection of malignancy

The AUC was 80% (95% confidence interval [CI] = 70.3–89.7; P < 0.001). At the cutoff value of ≥9.5, the sensitivity was 80%, specificity was 66.7%, LR+ was 2.40, and LR− was 0.30; at the cutoff value of 10.5, the sensitivity was 70.9%, specificity was 75%, LR+ was 2.84, and LR− was 0.0.39). For CD, the AUC was only 51.9% (P = 0.786) (Fig. 6).

Fig. 6.

AUC of Angio-PLUS and CD for the detection of malignancy using vascular imaging score. AUC, area under the receiver operating characteristic curve; CD, color Doppler.

The vascular patterns seen on Angio-PLUS showed a good correlation with histopathological results. The descriptor that showed a high predictive value for benignity was marginal orientation (90.5%), while the features most predictive of malignancy were radial orientation (96.9%) and mesh-like pattern (95.5%) (Table 3).

Table 3.

Correlation of vascular pattern descriptors with histopathology (n = 79).

Vascular pattern on color Doppler	Histopathology			Vascular pattern on Angio-PLUS	Histopathology
Vascular pattern on color Doppler	Benign (n = 24)	Malignant (n = 55)	P value	Vascular pattern on Angio-PLUS	Benign (n = 24)	Malignant (n = 55)	P value
No vascularity seen (n = 18)	6 (33.3)	12 (66.7)	0.757	No vascularity (n = 0)	–	–
External-dot-spot (n = 0)	–	–	–	External-dot-spot (n = 1)	1 (100)	0 (0)	0.303
Internal-dot-spot (n = 14)	3 (21.4)	11 (78.6)	0.425	Internal-dot-spot (n = 3)	2 (66.7)	1 (33.3)	0.217
Marginal (n = 18)	15 (83.3)	3 (16.7)	<0.001	Marginal (n = 21)	19 (90.5)	2 (9.5)	<0.001
Radial (n = 29)	0	29 (100)	<0.001	Radial (n = 32)	1 (3.1)	31 (96.9)	<0.001
Mesh (n = 0)	–	–	–	Mesh (n = 22)	1 (4.5)	21 (95.5)	<0.001

Values are given as n (%), unless otherwise indicated, and compared using Fisher's exact test. P < 0.05 is significant (in bold).

Discussion

Tumor vascularity plays a major role in the development and growth of tumors and is believed to indicate their biological behavior. The calculation of micro-vessel density (MVD) can provide information about tumor vascularity, which is routinely assessed on histopathology, and which is also considered the gold standard method for MVD. However, it requires a biopsy.

Advances in US techniques with improvements in the spatial resolution and speed of imaging have triggered a renewed interest in the vascularity assessment of tumors by US-based techniques. US-based MVI can be of two types, based on whether contrast media is used or not (16). Non-contrast MVI techniques utilize intelligent wall filtering systems to separate flow signals from clutter, thus removing only the clutter and preserving the slow flow signals, to allow visualization of more micro-vessels with high resolution (13 –15). The SMI utilizes a multi-dimensional wall filter to separate low-velocity flow signals from overlapping clutter artifacts (13 –15).

Angio-PLUS gives detailed information on microvascular flow by using unfocused or plane waves and three-dimensional (3D) wall filtering (26). A pulse repetition of unfocused or plane waves is sent into the body, and each pixel is interrogated continuously, with a 5–10 times faster sampling rate than CD (23). Blood flow information is extracted by using wall filtering. However, wall filtering cannot extract flows moving at varying speeds from tissue, such as those in small vessels. 3D wall filtering is used to analyze tissue motion in time, space, and amplitude domains, which effectively distinguishes between flow and tissue (27).

In this study, we used the Angio-PLUS technique along with conventional US methods to analyze tumor vascularity and internal vascular patterns.

We have described vascularity in two contexts in this study: (i) the vascular score, which is in the range of 0–13, as shown in Table 2; and (ii) the vascular pattern (Table 3) (not the score) of lesions.

Since this study was not conducted in a screening population, the majority of masses were >10 mm in size, and it was expected that vascularity would be detectable in most lesions, yet CD was unable to detect vessels in 18/79 (22.7%) lesions. When considering all lesions, Angio-PLUS offered better performance than CD, with a median vascular score of 11 (significantly higher than that on CD, which had a score of only 5 points).

Malignant lesions showed a significantly higher vascular score compared to benign lesions (median = 11 vs. 9; P < 0.001) on Angio-PLUS imaging. However, on CD there was no significant difference in the vascular scores of benign and malignant masses, indicating superiority of Angio-PLUS over CD. Moreover, CD was not sensitive enough to detect vascularity in 12/55 (21.8%) malignant lesions. This could be a significant drawback in settings where follow-up imaging is planned for response assessment (during neoadjuvant chemotherapy) for patients who cannot undergo magnetic resonance imaging scans or in resource-limited settings where US-based follow-ups are often used.

The AUC for Angio-PLUS was 0.800, which was statistically significant in differentiating benign versus malignant. The best cutoff value to differentiate benign from malignant lesions was ≥9.5; however, its sensitivity (80%) and specificity (66.75%) were not high enough to reliably differentiate benign from malignant masses.

Park et al. (21) calculated cutoff values of 5 with a sensitivity of 82.3% and a specificity of 65.3% (AUC = 0.808), which is comparable to the present study. The higher cutoff values in our study could be explained by the technical differences in the SMI and Angio-PLUS techniques. The mere presence of internal vascularity is not sufficient to indicate malignancy. The specificity of vascular scoring could be limited due to overlapping Doppler signs between hyper-vascular benign tumors, such as fibroadenomas or intraductal papillomas, hypo-vascular malignant tumors, such as ductal carcinoma in situ, and small invasive carcinomas (28). In this context, we would like to highlight that technical advancements in CD equipment would also lead to the detection of more vascularity in a higher number of benign lesions. Therefore, instead of focusing only on the mean and median vascular scores, there is a need to explore possible differences in the vascular patterns within lesions.

Several authors have used aspects of vascular morphology on CD, in differentiating benign from malignant lesions. Raza and Baum (29) described the positioning of vessels in and around lesions as being useful for predicting malignancy. They found that malignant lesions often showed prominent vessels entering the mass at its periphery (68%) and penetrating it (68%) with an irregular branching pattern. Benign lesions were predominantly avascular or sometimes showed small central vessels and vessels around the periphery. Weind et al. (9) suggested that evaluation of the spatial distribution of vessels and the distribution of peripheral microvascular density in carcinomas can be useful in the work-up of indeterminate breast lesions. They demonstrated that larger vessels (>40 μm) were found in the periphery of carcinomas compared to a more uniform distribution in fibroadenomas. Lee et al. (7) analyzed the location, shape, and penetration of Doppler signals and concluded that the presence of signals within a solid breast mass penetrating from the periphery into the center in a branching pattern strongly suggested malignancy.

Svensson et al. (25) categorized lesion vascularity into four major groups based on vessel distribution (peripheral, internal, mixed peripheral-internal, and spot vessels) and subcategorized them according to the vessel orientation (marginal, parallel, and radial).

We analyzed the vascular patterns of lesions qualitatively, using methods modified from those described initially by Svensson et al. (25). The basis of classification into five groups was vessel distribution and orientation and presence/absence of mesh pattern. This method yielded good results to differentiate benign from malignant lesions.

In our study, CD was able to identify radial vessels in malignant lesions but could not demonstrate the mesh pattern. The internal dot-spot pattern was seen in 14 cases (11 malignant, 3 benign) on CD, whereas only three cases (1 malignant, 2 benign) on Angio-PLUS showed this pattern, probably due to the high sensitivity of Angio-PLUS, which could detect smaller vessels with more clarity in most lesions (implying that Angio-PLUS was able to reveal larger segments of internal vessels in lesions whereas CD detected internal vessels as “dots and spots” only). The ability of Angio-PLUS to provide more exquisite details about vessels can be potentially useful in response assessment after neoadjuvant therapy.

The vascular patterns most predictive of malignancy in this study (on Angio-PLUS) were mesh (95.5%) and radial (96.9%) seen in 52/55 cases, whereas benign lesions showed a marginal orientation of vessels in 19/24 (90.5%) cases. The marginal orientation of vessels had a high negative predictive value of 90.5%. Only two of the benign lesions showed a radial or mesh pattern (i.e. two false-positives if we considered this pattern as a predictor of malignancy). However, on closer inspection of vessel details in these two cases, we were able to appreciate that the vessels were finer / more slender, smoother in outline compared to those seen in malignant masses. No significant vessel tortuosity or color-aliasing was noted in these two lesions, possibly implying that only lesion vascularity was increased without disordered angiogenesis. Since we had only two such cases, generalizability of this criteria could not be assessed. If more intricate vessel details are recorded, better lesion characterization may be possible and Angio-PLUS could have added value when deciding the radiology-pathology concordance of biopsied lesions.

The present study has some limitations. These include that dedicated measurements of vessel diameters and the presence/absence of color aliasing were not noted for all cases, and only a subjective visual comparison was available. Due to the relatively small sample size, these observations need to be interrogated in future studies. Moreover, the results can be affected by technical factors, such as equipment sensitivity settings; different institutions may evaluate vascularity differently. To overcome this limitation, our equipment had constant and optimal technical settings for all Doppler examinations, as described in the “Material and Methods” section. Even though we selected concise definitions for the features analyzed, the possibility of interpretation errors could exist; to minimize these, two experienced observers reviewed the images and a consensus opinion was taken. Since this study was conducted in a diagnostic environment, the lesion sizes were relatively large and the preliminary diagnostic information provided by grayscale features was also useful in predicting the likelihood of malignancy; however, in smaller lesions with indeterminate grayscale features, the integration of Angio-PLUS features may improve the diagnostic accuracy of imaging. Future studies in screening settings may be undertaken to interrogate whether the detection of tiny vessels within lesions or their pattern / morphology can influence a change in BI-RADS assessment. Based on this study, we cannot conclude that biopsies can be avoided; however, better decisions about the radiology-pathology concordance may be possible. In addition, a higher sensitivity of Angio-PLUS may be potentially useful during follow-up for response assessment during neoadjuvant therapy.

In conclusion, Angio-PLUS was more sensitive in depicting vascularity in breast masses than CD and superior in differentiating benign from malignant breast masses. However, specificity may be limited if only the vascular imaging scores are used. A qualitative analysis of vascular patterns (a combination of vessel distribution, orientation, and presence / absence of mesh pattern) on Angio-PLUS was found to be useful.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Namita Mohindra

Neeraj Jain

Prabhakar Mishra

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;71:209–249.

Weidner

Folkman

Pozza

, et al. Tumor angiogenesis: a new significant and independent prognostic indicator in early-stage breast carcinoma. J Natl Cancer Inst 1992;84:1875–1887.

Weidner

Semple

Welch

, et al. Tumor angiogenesis and metastasis—correlation in invasive breast carcinoma. N Engl J Med 1991;324:1–8.

Less

Skalak

Sevick

, et al. Microvascular architecture in a mammary carcinoma: branching patterns and vessel dimensions. Cancer Res 1991;51:265–273.

Mehta

Raza

Baum

. Use of Doppler ultrasound in the evaluation of breast carcinoma. Semin Ultrasound, CT and MRI 2000;21:297–307.

Sahin-Akyar

Sumer

. Color Doppler ultrasound and spectral analysis of tumor vessels in the differential diagnosis of solid breast masses. Invest Radiol 1996;31:72–79.

Lee

Choi

Baek

, et al. Role of color and power Doppler imaging in differentiating between malignant and benign solid breast masses. J Clin Ultrasound 2002;30:459–464.

Cosgrove

Kedar

Bamber

, et al. Breast diseases: color Doppler US in differential diagnosis. Radiology 1993;189:99–104.

Weind

Maier

Rutt

, et al. Invasive carcinomas and fibroadenomas of the breast: comparison of micro vessel distributions-implications for imaging modalities. Radiology 1998;208:477–483.

10.

Peters-Engl

Medl

Mirau

, et al. Color-coded and spectral Doppler flow in breast carcinomas–relationship with the tumor microvasculature. Breast Cancer Res Treat 1998;47:83–89.

11.

Schroeder

Bostanjoglo

Rademaker

, et al. Role of power Doppler techniques and ultrasound contrast enhancement in the differential diagnosis of focal breast lesions. Eur Radiol 2003;13:68–79.

12.

Buadu

Murakami

Murayama

, et al. Color Doppler sonography of breast masses: a multiparameter analysis. Clin Radiol 1997;52:917–923.

13.

Yoo

Choo

. Ultrasonographic demonstration of the tissue microvasculature in children: microvascular ultrasonography versus conventional color Doppler ultrasonography. Korean J Radiol 2020;21:146–158.

14.

Park

Seo

. Up-to-date Doppler techniques for breast tumor vascularity: superb microvascular imaging and contrast enhanced ultrasound. Ultrasonography 2018;37:98–106.

15.

Hata

. Seeing the unseen: new techniques in vascular imaging Superb Microvascular Imaging. Toshiba Med Rev 2014 Aug 13:1–8.

16.

Park

Seo

Han

. Breast ultrasound microvascular imaging and radiogenomics. Korean J Radiol 2021;22:677–687.

17.

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;20:759–772.

18.

Zhang

, et al. Vascular index measured by smart 3-D superb microvascular imaging can help to differentiate malignant and benign breast lesion. Cancer Manag Res 2019;11:5481–5487.

19.

Chae

Yoon

Cha

, et al. Added value of the vascular index on superb microvascular imaging for the evaluation of breast masses: comparison with grayscale ultrasound. J Ultrasound Med 2020;40:715–723.

20.

Kim

Lee

, et al. New Doppler imaging technique for assessing angiogenesis in breast tumors: correlation with immunohistochemically analyzed microvessels density. Acta Radiol 2018;59:1414–1421.

21.

Park

Seo

Woo

, et al. The utility of ultrasound superb microvascular imaging for evaluation of breast tumour vascularity: comparison with colour and power Doppler imaging regarding diagnostic performance. Clin Radiol 2018;73:304–311.

22.

, et al. The diagnostic value of superb microvascular imaging (SMI) in detecting blood flow signals of breast lesions: a preliminary study comparing SMI to color Doppler flow imaging. Medicine (Baltimore) 2015;94:e1502.

23.

Bercoff

Frappart

. White Paper. Angio PL.U.S. PLaneWave UltraSensitive™ ultrasound imaging. SuperSonic Imagine. 2016; 1–8.

24.

D’Orsi

Sickles

Mendelson

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

25.

Svensson

Pandian

Hashimoto

. The use of breast ultrasound color Doppler vascular pattern morphology improves diagnostic sensitivity with minimal change in specificity. Ultraschall Med 2010;31:466–474.

26.

Jung

Park

, et al. Comparison of the diagnostic performance of power Doppler ultrasound and a new microvascular Doppler ultrasound technique (AngioPLUS) for differentiating benign and malignant breast masses. J Ultrasound Med 2018;37:2689–2698.

27.

Bercoff

Tanter

. Ultrasound imaging goes ultrafast: a change in paradigm in medical ultrasound. Med Phys Int J 2015;3:109–119.

28.

Giuseppetti

Baldassarre

Marconi

. Color Doppler sonography. Eur J Radiol 1998;27:S254–S258.

29.

Raza

Baum

. Solid breast lesions: evaluation with power Doppler US. Radiology 1997;203:164–168.