A machine learning approach for differentiating malignant from benign enhancing foci on breast MRI
Tóm tắt
Differentiate malignant from benign enhancing foci on breast magnetic resonance imaging (MRI) through radiomic signature. Forty-five enhancing foci in 45 patients were included in this retrospective study, with needle biopsy or imaging follow-up serving as a reference standard. There were 12 malignant and 33 benign lesions. Eight benign lesions confirmed by over 5-year negative follow-up and 15 malignant histopathologically confirmed lesions were added to the dataset to provide reference cases to the machine learning analysis. All MRI examinations were performed with a 1.5-T scanner. One three-dimensional T1-weighted unenhanced sequence was acquired, followed by four dynamic sequences after intravenous injection of 0.1 mmol/kg of gadobenate dimeglumine. Enhancing foci were segmented by an expert breast radiologist, over 200 radiomic features were extracted, and an evolutionary machine learning method (“training with input selection and testing”) was applied. For each classifier, sensitivity, specificity and accuracy were calculated as point estimates and 95% confidence intervals (CIs). A k-nearest neighbour classifier based on 35 selected features was identified as the best performing machine learning approach. Considering both the 45 enhancing foci and the 23 additional cases, this classifier showed a sensitivity of 27/27 (100%, 95% CI 87–100%), a specificity of 37/41 (90%, 95% CI 77–97%), and an accuracy of 64/68 (94%, 95% CI 86–98%). This preliminary study showed the feasibility of a radiomic approach for the characterisation of enhancing foci on breast MRI.
Tài liệu tham khảo
Lee-Felker SA, Tekchandani L, Thomas M et al (2017) Newly diagnosed breast cancer: comparison of contrast-enhanced spectral mammography and breast MR imaging in the Evaluation of Extent of Disease. Radiology 285:389–400. https://doi.org/10.1148/radiol.2017161592
Fallenberg E, Dromain C, Diekmann F et al (2014) Contrast-enhanced spectral mammography versus MRI: initial results in the detection of breast cancer and assessment of tumour size. Eur Radiol 24:256–264. https://doi.org/10.1007/s00330-013-3007-7
Fallenberg EM, Schmitzberger FF, Amer H et al (2017) Contrast-enhanced spectral mammography vs. mammography and MRI--clinical performance in a multi-reader evaluation. Eur Radiol 27:2752–2764. https://doi.org/10.1007/s00330-016-4650-6
Mauri G, Sconfienza LM, Pescatori LC et al (2017) Technical success, technique efficacy and complications of minimally-invasive imaging-guided percutaneous ablation procedures of breast cancer: a systematic review and meta-analysis. Eur Radiol 27:3199–3210. https://doi.org/10.1007/s00330-016-4668-9
Lewis TC, Pizzitola VJ, Giurescu ME et al (2017) Contrast-enhanced digital mammography: a single-institution experience of the first 208 cases. Breast J 23:67–76. https://doi.org/10.1111/tbj.12681
Ali-Mucheru M, Pockaj B, Patel B et al (2016) Contrast-enhanced digital mammography in the surgical management of breast cancer. Ann Surg Oncol 23:649–655. https://doi.org/10.1245/s10434-016-5567-7
Clauser P, Cassano E, De Nicolò A et al (2016) Foci on breast magnetic resonance imaging in high-risk women: cancer or not? Radiol Med 121:611–617. https://doi.org/10.1007/s11547-016-0644-3
Cheng L, Li X (2012) Breast magnetic resonance imaging: focus/foci. Gland Surg 1:136–138. https://doi.org/10.3978/j.issn.2227-684X.2012.08.06
Ha R, Sung J, Lee C, Comstock C, Wynn R, Morris E (2014) Characteristics and outcome of enhancing foci followed on breast MRI with management implications. Clin Radiol 69:715–720. https://doi.org/10.1016/j.crad.2014.02.007
Liberman L, Mason G, Morris EA, Dershaw DD (2006) Does size matter? Positive predictive value of MRI-detected breast lesions as a function of lesion size. AJR Am J Roentgenol 186:426–430. https://doi.org/10.2214/AJR.04.1707
Bahrs SD, Baur A, Hattermann V et al (1987) (2014) BI-RADS® 3 lesions at contrast-enhanced breast MRI: is an initial short-interval follow-up necessary? Acta Radiol 55:260–265. https://doi.org/10.1177/0284185113501304
Jansen SA, Shimauchi A, Zak L, Fan X, Karczmar GS, Newstead GM (2011) The diverse pathology and kinetics of mass, nonmass, and focus enhancement on MR imaging of the breast. J Magn Reson Imaging 33:1382–1389. https://doi.org/10.1002/jmri.22567
Sardanelli F, Boetes C, Borisch B et al (2010) Magnetic resonance imaging of the breast: recommendations from the EUSOMA working group. Eur J Cancer 46:1296–1316. https://doi.org/10.1016/j.ejca.2010.02.015
Lambin P, Rios-Velazquez E, Leijenaar R et al (2012) Radiomics: extracting more information from medical images using advanced feature analysis. Eur J Cancer 48:441–446. https://doi.org/10.1016/j.ejca.2011.11.036
Bland K, Copeland E, Klimberg VS, Gradishar W (2018) The Breast - 5th Edition. https://www.elsevier.com/books/the-breast/bland/978-0-323-35955-9. Accessed 6 Mar 2018
Gillies RJ, Kinahan PE, Hricak H (2015) Radiomics: images are more than pictures, they are data. Radiology 278:563–577. https://doi.org/10.1148/radiol.2015151169
ACR BI-RADS Atlas® 5th Edition https://www.acr.org/Clinical-Resources/Reporting-and-Data-Systems/Bi-Rads. Accessed 6 Mar 2019
(2018) StackReg. Available via http://bigwww.epfl.ch/thevenaz/stackreg/. Accessed 6 Mar 2018
(2018) Image J. Available via https://imagej.nih.gov/ij/. Accessed 6 Mar 2018
(2018) ITK - Segmentation & Registration Toolkit. Available via https://itk.org/. Accessed 6 Mar 2018
(2018) ITK: itk::LabelGeometryImageFilter< TLabelImage, TIntensityImage > Class Template Reference. Available via https://itk.org/Doxygen/html/classitk_1_1LabelGeometryImageFilter.html. Accessed 6 Mar 2018
Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A (2018) Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. https://doi.org/10.3322/caac.21492
(2018) Female Breast Cancer - Cancer Stat Facts. Available via https://seer.cancer.gov/statfacts/html/breast.html. Accessed 12 Mar 2018
Blumen H, Fitch K, Polkus V (2016) Comparison of treatment costs for breast cancer, by tumor stage and type of service. Am Health Drug Benefits 9:23–32
Buscema M, Breda M, Lodwick W (2013) Training With Input Selection and Testing (TWIST) algorithm: a significant advance in pattern recognition performance of machine learning. J Intell Learn Syst Appl 5:29. https://doi.org/10.4236/jilsa.2013.51004