Abstract
Purpose
Early diagnosis of triple-negative (TN) breast cancer is important due to its aggressive biological characteristics, poor clinical outcomes, and limited options for therapy. The goal of this study is to evaluate the potential of machine learning with quantitative ultrasound image features for the diagnosis of TN breast cancer.
Methods
Ultrasonic and clinical data of 140 surgically confirmed breast cancer cases were analyzed retrospectively for the diagnosis of TN and non-TN (NTN) subtypes. The subtypes were classified based on the expression of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2). Ultrasound image features were measured from the grayscale and color Doppler images and used with logistic regression for classification by machine learning. Leave-one-out cross validation was used to train and test the differentiation. Diagnostic performance was measured by the area under receiver operating characteristic (ROC) curve, and sensitivity and specificity determined at the Youdons index.
Results
Of the twelve grayscale and Doppler features measured, eight were found to be statistically different for the TN and NTN subtypes (p < 0.05). The area under the ROC curve (AUC) of the statistically significant grayscale (GS) and color Doppler (CD) features was 0.85 and 0.65, respectively. The AUC increased to 0.88 when the GS and CD features were used together, with sensitivity of 86.96% and specificity of 82.91%. Consideration of patient age in the analysis did not improve discrimination of TN and NTN.
Conclusions
The analysis of breast ultrasound images by machine learning achieves high level of differentiation between the TN and NTN subtypes, exceeding the diagnostic performance by standard visual assessments of the images.
Similar content being viewed by others
References
Torre LA, Bray F, Siegel RL et al (2015) Global cancer statistics, 2012. CA Cancer J Clin 65:87–108. https://doi.org/10.3322/caac.21262
Siegel RL, Miller KD, Jemal A (2018) Cancer statistics, 2018. CA Cancer J Clin 68:7–30. https://doi.org/10.3322/caac.21442
Lin NU, Vanderplas A, Hughes ME et al (2012) Clinicopathologic features, patterns of recurrence, and survival among women with triple-negative breast cancer in the National Comprehensive Cancer Network. Cancer 118:5463–5472. https://doi.org/10.1002/cncr.27581
Sasaki Y, Tsuda H (2009) Clinicopathological characteristics of triple-negative breast cancers. Breast Cancer 16:254–259. https://doi.org/10.1007/s12282-009-0153-5
Stavros AT, Thickman D, Rapp CL et al (1995) Solid breast nodules: use of sonography to distinguish between benign and malignant lesions. Radiology 196:123–134. https://doi.org/10.1148/radiology.196.1.7784555
Sehgal CM, Weinstein SP, Arger PH, Conant EF (2006) A Review of Breast Ultrasound. J Mammary Gland Biol Neoplasia 11:113–123. https://doi.org/10.1007/s10911-006-9018-0
Ko ES, Lee BH, Kim H-A et al (2010) Triple-negative breast cancer: correlation between imaging and pathological findings. Eur Radiol 20:1111–1117. https://doi.org/10.1007/s00330-009-1656-3
Krizmanich-Conniff KM, Paramagul C, Patterson SK et al (2012) Triple receptor–negative breast cancer: imaging and clinical characteristics. Am J Roentgenol 199:458–464
Boisserie-Lacroix M, Macgrogan G, Debled M et al (2013) Triple-negative breast cancers: associations between imaging and pathological findings for triple-negative tumors compared with hormone receptor-positive/human epidermal growth factor receptor-2-negative breast cancers. Oncologist 18:802–811. https://doi.org/10.1634/theoncologist.2013-0380
Yassin NIR, Omran S, El Houby EMF, Allam H (2018) Machine learning techniques for breast cancer computer aided diagnosis using different image modalities: a systematic review. Comput Methods Programs Biomed 156:25–45. https://doi.org/10.1016/j.cmpb.2017.12.012
Sehgal CM, Cary TW, Kangas SA et al (2004) Computer-based margin analysis of breast sonography for differentiating malignant and benign masses. J Ultrasound Med 23:1201–1209
Huang Y-S, Takada E, Konno S et al (2018) Computer-Aided tumor diagnosis in 3-D breast elastography. Comput Methods Progr Biomed 153:201–209. https://doi.org/10.1016/j.cmpb.2017.10.021
Bouzghar G, Levenback BJ, Sultan LR et al (2014) Bayesian probability of malignancy with BI-RADS sonographic features. J Ultrasound Med 33:641–648. https://doi.org/10.7863/ultra.33.4.641
Breast Imaging Reporting Data System | American College of Radiology. https://www.acr.org/Clinical-Resources/Reporting-and-Data-Systems/Bi-Rads#Ultrasound. Accessed 23 Jul 2018
Goldhirsch A, Winer EP, Coates AS et al (2013) Personalizing the treatment of women with early breast cancer: highlights of the St Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2013. Ann Oncol 24:2206–2223. https://doi.org/10.1093/annonc/mdt303
Sehgal CM, Cary TW, Cwanger A et al (2012) Combined Naïve Bayes and logistic regression for quantitative breast sonography. In: 2012 IEEE International Ultrasonics Symposium. IEEE, pp 1686–1689
Sultan LR, Xiong H, Zafar HM et al (2015) Vascularity assessment of thyroid nodules by quantitative color Doppler ultrasound. Ultrasound Med Biol 41:1287–1293. https://doi.org/10.1016/j.ultrasmedbio.2015.01.001
Prat A, Perou CM (2011) Deconstructing the molecular portraits of breast cancer. Mol Oncol 5:5–23. https://doi.org/10.1016/j.molonc.2010.11.003
Dairkee SH, Ljung BM, Smith H, Hackett A (1987) Immunolocalization of a human basal epithelium specific keratin in benign and malignant breast disease. Breast Cancer Res Treat 10:11–20
Carey LA, Dees EC, Sawyer L et al (2007) The triple negative paradox: primary tumor chemosensitivity of breast cancer subtypes. Clin Cancer Res 13:2329–2334. https://doi.org/10.1158/1078-0432.CCR-06-1109
Wojcinski S, Soliman AA, Schmidt J et al (2012) Sonographic features of triple-negative and non-triple-negative breast cancer. J Ultrasound Med 31(10):1531–1541
Tang J, Rangayyan RM, Xu J et al (2009) Computer-aided detection and diagnosis of breast cancer with mammography: recent advances. IEEE Trans Inf Technol Biomed 13:236–251. https://doi.org/10.1109/TITB.2008.2009441
Dromain C, Boyer B, Ferré R et al (2013) Computed-aided diagnosis (CAD) in the detection of breast cancer. Eur J Radiol 82:417–423. https://doi.org/10.1016/j.ejrad.2012.03.005
Xiong H, Sultan LR, Cary TW et al (2017) The diagnostic performance of leak-plugging automated segmentation versus manual tracing of breast lesions on ultrasound images. Ultrasound 25:98–106. https://doi.org/10.1177/1742271X17690425
Dogan BE, Turnbull LW (2012) Imaging of triple-negative breast cancer. Ann Oncol 23(Suppl 6):vi23–29. https://doi.org/10.1093/annonc/mds191
Shin HJ, Kim HH, Huh MO et al (2011) Correlation between mammographic and sonographic findings and prognostic factors in patients with node-negative invasive breast cancer. Br J Radiol 84:19–30. https://doi.org/10.1259/bjr/92960562
Yang W-T, Dryden M, Broglio K et al (2008) Mammographic features of triple receptor-negative primary breast cancers in young premenopausal women. Breast Cancer Res Treat 111:405–410. https://doi.org/10.1007/s10549-007-9810-6
Elkabets M, Gifford AM, Scheel C et al (2011) Human tumors instigate granulin-expressing hematopoietic cells that promote malignancy by activating stromal fibroblasts in mice. J Clin Invest 121:784–799. https://doi.org/10.1172/JCI43757
Li J-W, Zhang K, Shi Z-T et al (2018) Triple-negative invasive breast carcinoma: the association between the sonographic appearances with clinicopathological feature. Sci Rep 8:9040. https://doi.org/10.1038/s41598-018-27222-6
Yoon GY, Cha JH, Kim HH et al (2018) Sonographic features that can be used to differentiate between small triple-negative breast cancer and fibroadenoma. Ultrasonography 37:149–156. https://doi.org/10.14366/usg.17036
Costantini M, Belli P, Bufi E et al (2016) Association between sonographic appearances of breast cancers and their histopathologic features and biomarkers. J Clin Ultrasound 44:26–33. https://doi.org/10.1002/jcu.22312
Schrading S, Kuhl CK (2008) Mammographic, US, and MR imaging phenotypes of familial breast cancer. Radiology 246:58–70. https://doi.org/10.1148/radiol.2461062173
Verheul HMW, Voest EE, Schlingemann RO (2004) Are tumours angiogenesis-dependent? J Pathol 202:5–13. https://doi.org/10.1002/path.1473
Sehgal CM, Arger PH, Rowling SE et al (2000) Quantitative vascularity of breast masses by Doppler imaging: regional variations and diagnostic implications. J Ultrasound Med 19:427 – 40; quiz 441–2
Zhang L, Li J, Xiao Y et al (2015) Identifying ultrasound and clinical features of breast cancer molecular subtypes by ensemble decision. Sci Rep 5:11085. https://doi.org/10.1038/srep11085
Kumar R, Yarmand-Bagheri R (2001) The role of HER2 in angiogenesis. Semin Oncol 28:27–32
Dogan BE, Gonzalez-Angulo AM, Gilcrease M et al (2010) Multimodality imaging of triple receptor-negative tumors with mammography, ultrasound, and MRI. AJR Am J Roentgenol 194:1160–1166. https://doi.org/10.2214/AJR.09.2355
Kojima Y, Tsunoda H (2011) Mammography and ultrasound features of triple-negative breast cancer. Breast Cancer 18:146–151. https://doi.org/10.1007/s12282-010-0223-8
Lerma E, Peiro G, Ramón T et al (2007) Immunohistochemical heterogeneity of breast carcinomas negative for estrogen receptors, progesterone receptors and Her2/neu (basal-like breast carcinomas). Mod Pathol 20:1200–1207. https://doi.org/10.1038/modpathol.3800961
Yeo SH, Kim GR, Lee SH, Moon WK (2018) Comparison of ultrasound elastography and color Doppler ultrasonography for distinguishing small triple-negative breast cancer from fibroadenoma. J Ultrasound Med. https://doi.org/10.1002/jum.14564
Yang Q, Liu H-Y, Liu D, Song Y-Q (2015) Ultrasonographic features of triple-negative breast cancer: a comparison with other breast cancer subtypes. Asian Pac J Cancer Prev 16:3229–3232
Du H-Y, Lin B-R, Huang D-P (2015) Ultrasonographic findings of triple-negative breast cancer. Int J Clin Exp Med 8:10040–10043
Bauer KR, Brown M, Cress RD et al (2007) Descriptive analysis of estrogen receptor (ER)-negative, progesterone receptor (PR)-negative, and HER2-negative invasive breast cancer, the so-called triple-negative phenotype. Cancer 109:1721–1728. https://doi.org/10.1002/cncr.22618
Acknowledgements
This research was supported by NIH grant RO1 CA130946 and by the National Natural Science Foundation of China grant 81630048, 81271647.
Author information
Authors and Affiliations
Corresponding authors
Ethics declarations
Conflict of interest
All authors declare they have no conflict of interest.
Ethical approval
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent
Informed consent was obtained from all individual participants included in the study.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Rights and permissions
About this article
Cite this article
Wu, T., Sultan, L.R., Tian, J. et al. Machine learning for diagnostic ultrasound of triple-negative breast cancer. Breast Cancer Res Treat 173, 365–373 (2019). https://doi.org/10.1007/s10549-018-4984-7
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10549-018-4984-7