Radiologists’ performance in the detection of benign and malignant masses with 3D automated breast ultrasound (ABUS)

https://doi.org/10.1016/j.ejrad.2011.01.074Get rights and content

Abstract

Objectives

To retrospectively evaluate the detection performance of benign and malignant breast masses using 3D volume data obtained by ABUS and to determine lesion variables which affect detectability.

Methods

Between November and December of 2007, bilateral whole breast US images were obtained using ABUS in 67 consecutive women who were scheduled to undergo US-guided needle biopsy due to suspicious breast masses. Twenty-four invasive ductal cancers in 23 breasts, 46 benign breast lesions in 44 breasts and 38 normal breasts were included. Three breast radiologists (experience range, 8–16 years) who did not perform the examinations and were blinded to the histology independently reviewed the ABUS data of the 105 breasts to detect suspicious solid masses with pathology as the standard of reference. Sensitivity and specificity in detecting benign and malignant masses were calculated, and lesion characteristics affecting detectability were analyzed.

Results

Sensitivities for benign and malignant mass detections were 65.2% (30/46), 95.8% (23/24) for reader 1 (p = 0.007), 66.7% (31/46), 87.5% (21/24) for reader 2 (p = 0.087), and 56.3% (24/46), 91.7% (22/24), for reader 3 (p = 0.001), respectively. Logistic analysis showed that mass size (odds ratio, 95% CI; 1.12, 1.02–1.24), surrounding tissue changes (odds ratio, 95% CI; 0.11, 0.02–0.47), and shape of the mass (odds ratio, 95% CI; 3.12, 1.02–9.55) were the variables associated with detectability at ABUS.

Conclusion

In reader studies using ABUS data, significantly higher sensitivity was noted for malignant breast masses than for benign masses.

Introduction

Screening mammography is accepted as the only means of reducing breast cancer mortality [1], [2], but the reduced sensitivity of mammographic screening for women with dense breasts and for women younger than 50 years of age remains a major limitation [3], [4]. Recent research from the American College of Radiology Imaging Network (ACRIN) 6666 trial, which is the largest of its kind using screening ultrasound (US) with standardized scanning and interpretive criteria, has revealed that the addition of a single US examination to mammography for the screening of women who are at elevated risk of breast cancer can increase the detection of small, node-negative cancers [5]; the supplemental yield in their study was 4.2 per 1000 women screened, however, the time and skill necessary to detect small, non-palpable tumors using hand-held imaging and the risk of false-positive results have thus far discouraged its widespread use [6], [7], [8].

Automated breast US (ABUS) systems have been proposed as suitable for screening for breast cancer [9], [10], [11], [12], [13], [14], [15] and the current ABUS units, which are equipped with high-frequency broadband transducers, have the advantages of reproducibility, utility (in terms of surveying large areas of the breast), and reduced operator dependence compared to hand-held US (HHUS) devices [16]. Indeed, several studies have concluded that the performances of ABUS and HHUS are equivalent [17], [18], [19], [20], however, the detection performance of ABUS by radiologists simulating real clinical practice has not been adequately evaluated, and factors that may affect detectability have not yet been established.

The purpose of the current study was to retrospectively evaluate the detection performance of ABUS using 3D volume data for benign and malignant breast masses, and to find the lesion variables which affect the detectability.

Section snippets

Patients

This study was approved by the Institutional Review Board of our hospital, and informed consent was obtained from all patients. From November to December of 2007, two trained mammography technologists obtained whole breast US images of 134 breasts in 67 consecutive women (mean age, 47 ± 14 years; range, 20–79 years) who were scheduled to undergo US guided needle biopsy due to suspicious breast masses using a commercially available ABUS unit (SomoVu Scan Station; U-system, Inc., San Jose, CA).

Imaging findings and tumor detection

Lesion characteristics of both benign and malignancy masses are noted in Table 1. Mean lesion size of cancer on US was 15.5 mm (range, 4–22 mm) and that of benign lesions was 13.5 mm (range 5–37 mm). Most of US lesion characteristics were significant different between benign and malignant masses. Benign masses were categorized as BI-RADS 3 in 7 and BI-RADS 4 in 39. Malignant masses were categorized as BI-RADS 4 in 8 and BI-RADS 5 in 16.

Distribution of the patients’ and lesions’ characteristics

Discussion

In our study, the estimated overall sensitivity and specificity of the three radiologists for the detection of solid breast masses were 71.9% and 79.5%, respectively, using 3D ABUS data. For all three readers, the sensitivities for malignant masses (87.5–95.8%) were much higher than the sensitivities for benign masses (56.3–66.7%) using ABUS data. Detectability of the lesions was found to be associated with mass size, surrounding tissue changes, and shape of the mass on multivariable logistic

Conclusion

The automated 3D breast scanner is a promising diagnostic tool, and it showed higher performance in the detection of malignant, large, irregular shaped masses with surrounding changes, than benign, small, or round/oval masses without surrounding changes. More studies on the performance of ABUS for screening population should be performed to make ABUS as an adjunct to mammography.

Acknowledgement

The research was supported by the Converging Research Center Program through the Ministry of Education, Science and Technology (2010K001113).

References (24)

  • D.B. Kopans

    Breast cancer screening with ultrasonography

    Lancet

    (1999)
  • Y.H. Chou et al.

    Automated full-field breast ultrasonography: the past and the present

    J Med Ultrasound

    (2007)
  • D. Kotsianos-Hermle et al.

    Analysis of 107 breast lesions with automated 3D ultrasound and comparison with mammography and manual ultrasound

    Eur J Radiol

    (2009)
  • L.L. Humphrey et al.

    Breast cancer screening: a summary of the evidence for the U.S. Preventive Services Task Force

    Ann Intern Med

    (2002)
  • S.W. Duffy et al.

    The impact of organized mammography service screening on breast carcinoma mortality in seven Swedish counties

    Cancer

    (2002)
  • M.T. Mandelson et al.

    Breast density as a predictor of mammographic detection: comparison of interval- and screen-detected cancers

    J Natl Cancer Inst

    (2000)
  • T.M. Kolb et al.

    Comparison of the performance of screening mammography, physical examination, and breast US and evaluation of factors that influence them: an analysis of 27,825 patient evaluations

    Radiology

    (2002)
  • W.A. Berg et al.

    Combined screening with ultrasound and mammography vs mammography alone in women at elevated risk of breast cancer

    JAMA

    (2008)
  • W. Buchberger et al.

    Incidental findings on sonography of the breast: clinical significance and diagnostic workup

    Am J Roentgenol

    (1999)
  • W.A. Berg

    Tailored supplemental screening for breast cancer: what now and what next?

    AJR Am J Roentgenol

    (2009)
  • A.P. Harper et al.

    Ultrasound in the evaluation of solid breast masses

    Radiology

    (1983)
  • V.G. Maturo et al.

    Ultrasound of the whole breast utilizing a dedicated automated breast scanner

    Radiology

    (1980)
  • Cited by (56)

    • Three-dimensional automated breast ultrasound: Technical aspects and first results

      2019, Diagnostic and Interventional Imaging
      Citation Excerpt :

      A factor contributing to the improved detection rates is the additional information obtained by 3D ABUS on the coronal plane as lesions may be identified only in one of the three orthogonal planes [11,52]. The factors that could affect detectability at 3D ABUS are the size, shape and location of the lesion, the composition of breast parenchyma, surrounding tissue changes and technical parameters [43,53]. Chou et al. examined 182 lesions (47 carcinomas and 135 benign lesions) and presented a sensitivity equal to 98% and a specificity equal to 85% [54].

    • Automatic Breast Volume Scanner versus Handheld Ultrasound in Differentiation of Benign and Malignant Breast Lesions: A Systematic Review and Meta-analysis

      2019, Ultrasound in Medicine and Biology
      Citation Excerpt :

      Meanwhile, numerous studies have addressed application of the ABVS in diagnostic settings over the past decade (Grady et al. 2017; Van Zelst et al. 2015; Wang et al. 2016; Xu et al. 2014; Zheng et al. 2015). Some studies with relatively small patient populations have reported promising results (Chang et al. 2011; Meng et al. 2015; Wang et al. 2016; Wojcinski et al. 2013). In clinical practice, however, most radiologists now use conventional HHUS in the characterization and differentiation of breast lesions discovered through mammography or other screening technologies.

    • Automated Breast Ultrasonography (ABUS) in the Screening and Diagnostic Setting: Indications and Practical Use

      2018, Academic Radiology
      Citation Excerpt :

      Several studies emphasized the importance of breast coverage: the peripheral location of the lesions may reduce the diagnostic performance of ABUS compared to HHUS, especially in large breasts, and could be a reason for misdiagnosis of cancer (41). The mean diameter of the lesion is an important factor associated to lesion detectability for ABUS, as confirmed by several studies (24,29). In particular, the detection rate increases with lesion size.

    • Evaluation of an automated breast volume scanner according to the fifth edition of BI-RADS for breast ultrasound compared with hand-held ultrasound

      2018, European Journal of Radiology
      Citation Excerpt :

      Although the sensitivity of ABVS according to the BI-RADS descriptors used in our study was lower than that of HHUS from previous studies that were based on the fourth edition, we found no significant difference in sensitivity between ABVS and HHUS [13,21]. Previous ABVS studies based on the fourth BI-RADS edition reported a sensitivity of 71.9–96.5%, a specificity of 51.2–92.3%, and an AUC of 0.730–0.967 [1,13,20,28,29]. Previous studies based on the fifth BI-RADS edition reported the sensitivity of 97.1% for both ABVS and HHUS, the specificity of 27.8% for ABVS and 30.4% for HHUS, which there was no significant difference between ABVS and HHUS [30].

    View all citing articles on Scopus
    View full text