Segmentation of pectoral muscle from digital mammograms with depth-first search algorithm towards breast density classification

• Autorzy: Pawar Shivaji D. , Sharma Kamal Kr. , Sapate Suhas G. , Yadav Geetanjali Y.

Identyfikatory

DOI

10.1016/j.bbe.2021.08.005

• Języki publikacji: EN

Abstrakty

EN

Digital mammography acts as a unique screening technology to protect the lives of females against breast cancer for the past few decades. Mammographic breast density is a well-known biomarker and plays a substantial role in breast cancer prediction and treatments. Breast density is calculated based on the opacity of fibro-glandular tissue reflected on digital mammograms concerning the whole area of the breast. The opacity of pectoral muscle and fibro-glandular tissue is similar to each other; hence, the small presence of the pectoral muscle in the breast area can hamper the accuracy of breast density classification. Successful removal of pectoral muscle is challenging due to changes in shape, size, and texture of pectoral muscle in every MLO and LMO views of mammogram. In this article, the depth-first search (DFS) algorithm is proposed to remove artifacts and pectoral muscle from digital mammograms. In the proposed algorithm, image enhancement is performed to improve the pixel quality of the input image. The whole breast as a single connected component is identified from the background region to remove the artifacts and tags. The depth-first search method with and without the heuristic approach is used to delineate the pectoral muscle, and then final suppression is performed on it. This algorithm is tested on 2675 images of the DDSM dataset, which is further divided into four density classes as per BIRADs classification. Segmentation results are calculated individually on each BIRADs density class of the DDSM dataset. Results are validated subjectively by the expert’s Radiologist’s ground truth with segmentation accuracy and objectively by the Jaccard coefficient and a dice similarity coefficient. This algorithm is found robust on each density class and provides overall segmentation accuracy of 86.18%, a mean value of Jaccard index, and a Dice similarity coefficient of 0.9315 and 0.9548, respectively. The experimental results show that the proposed algorithms applied for pectoral muscle removal follow the ground truth marked by an expert radiologist. The proposed algorithm can be part of the pre-processing unit of breast density measurement and breast cancer detection system used during clinical practice.

Słowa kluczowe

EN

breast cancer; depth-first search; mammographic breast density; pre-processing; pectoral muscle removal

PL

nowotwór piersi; mammograficzna gęstość piersi; przetwarzanie wstępne

• Wydawca: Nałęcz Institute of Biocybernetics and Biomedical Engineering of the Polish Academy of Sciences ; Elsevier
• Czasopismo: Biocybernetics and Biomedical Engineering
• Rocznik: 2021
• Tom: Vol. 41, no. 3
• Strony: 1224--1241
• Opis fizyczny: Bibliogr. 49 poz., rys., tab., wykr.

Twórcy

autor

Pawar Shivaji D.

• Department of Computer Science and Engineering, Lovely Professional University, Jalandhar (P.B.), India

autor

Sharma Kamal Kr.

• School of Electronics and Electrical Engineering, Lovely Professional University, Jalandhar (P.B.), India

autor

Sapate Suhas G.

• Department of Computer Science and Engineering, Annasaheb Dange College of Engineering and Technology, Ashta, Sangli (M.H.), India

autor

Yadav Geetanjali Y.

• Lifeline Hospital Musaffah Abu Dhabi, United Arab Emirates

Bibliografia

• [1] Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2021. CA Cancer J Clin 2021;71:7–33. https://doi.org/10.3322/caac.v71.110.3322/caac.21654.
• [2] Destounis SV, Santacroce A, Arieno A. Update on breast density, risk estimation, and supplemental screening. Am J Roentgenol 2020;214:296–305. https://doi.org/10.2214/AJR.19.21994.
• [3] Sharma KK, Pawar SD, Bali B. Proactive preventive and evidence-based artificial intelligene models: future healthcare 2020:463–72. https://doi.org/10.1007/978-981-15-0633-8_44.
• [4] Sapate S, Talbar S, Mahajan A, Sable N, Desai S, Thakur M. Breast cancer diagnosis using abnormalities on ipsilateral views of digital mammograms. Biocybern Biomed Eng 2020;40:290–305. https://doi.org/10.1016/j.bbe.2019.04.008.
• [5] Ilesanmi AE, Chaumrattanakul U, Makhanov SS. Methods for the segmentation and classification of breast ultrasound images: a review. J Ultrasound 2021. https://doi.org/10.1007/s40477-020-00557-5.
• [6] Yap MH, Goyal M, Osman F, Martí R, Denton E, Juette A, et al. Breast ultrasound region of Interest detection and lesion localisation. Artif Intell Med 2020; 107. https://doi.org/10.1016/j.artmed.2020.101880.
• [7] Brandt KR, Scott CG, Miglioretti DL, Jensen MR, Mahmoudzadeh AP, Hruska C, et al. Automated volumetric breast density measures: differential change between breasts in women with and without breast cancer. Breast Cancer Res 2019;21. https://doi.org/10.1186/s13058-019-1198-9.
• [8] García HAV, Gotay CC, Wilson CM, Lohrisch CA, Lai AS, Aronson KJ, et al. Mammographic density parameters and breast cancer tumor characteristics among postmenopausal women. Breast Cancer Targets Ther 2019;11:261–71. https://doi.org/10.2147/BCTT.S192766.
• [9] Sapate SG, Mahajan A, Talbar SN, Sable N, Desai S, Thakur M. Radiomics based detection and characterization of suspicious lesions on full field digital mammograms. Comput Methods Programs Biomed 2018;163:1–20. https://doi.org/10.1016/j.cmpb.2018.05.017.
• [10] Moon WK, Lo CM, Chang JM, Huang CS, Chen JH, Chang RF. Quantitative ultrasound analysis for classification of BI-RADS category 3 breast masses. J Digit Imaging 2013;26:1091–8. https://doi.org/10.1007/s10278-013-9593-8.
• [11] Caballo M, Pangallo DR, SanderinkW, Hernandez AM, Lyu SH, Molinari F, et al. Multi-marker quantitative radiomics for mass characterization in dedicated breast C.T. imaging. Med Phys 2021;48:313–28. https://doi.org/10.1002/mp.v48.110.1002/mp.14610.
• [12] Pawar S, Sapate S, Sharma K. Machine learning approach towards mammographic breast density measurement for breast cancer risk prediction: an overview. SSRN Electron J 2020. https://doi.org/10.2139/ssrn.3599187.
• [13] Pour SG, McLeod P, Verma B, Maeder A. Comparing data mining with ensemble classification of breast cancer masses in digital mammograms. CEUR Workshop Proc 2012;941:55–63.
• [14] Phipps AI, Buist DSM, Malone KE, Barlow WE, Porter PL, Kerlikowske K, et al. Breast density, body mass index, and risk of tumor marker-defined subtypes of breast cancer. Ann Epidemiol 2012;22:340–8. https://doi.org/10.1016/j.annepidem.2012.02.002.
• [15] Caballo M, Pangallo DR, Mann RM, Sechopoulos I. Deep learning-based segmentation of breast masses in dedicated breast C.T. imaging: radiomic feature stability between radiologists and artificial intelligence. Comput Biol Med 2020;118:103629. https://doi.org/10.1016/j.compbiomed.2020.103629.
• [16] Goudarzi M, Maghooli K. Extraction of fuzzy rules at different concept levels related to image features of mammography for diagnosis of breast cancer. Biocybern Biomed Eng 2018;38:1004–14. https://doi.org/10.1016/j.bbe.2018.09.002.
• [17] Sabeena Beevi K, Nair MS, Bindu GR. Automatic mitosis detection in breast histopathology images using Convolutional Neural Network based deep transfer learning. Biocybern Biomed Eng 2019;39:214–23. https://doi.org/10.1016/j.bbe.2018.10.007.
• [18] Shahid AH, Singh MP. Computational intelligence techniques for medical diagnosis and prognosis: problems and current developments. Biocybern Biomed Eng. 2019;39:638–72. https://doi.org/10.1016/j.bbe.2019.05.010.
• [19] Sapate S, Talbar S. An overview of pectoral muscle extraction algorithms applied to digital mammograms. vol. 651. 2016. https://doi.org/10.1007/978-3-319-33793-7_2.
• [20] Saltanat N, Hossain MA, Alam MS. An efficient pixel value based mapping scheme to delineate pectoral muscle from mammograms. Proc 2010 IEEE 5th Int Conf Bio-Inspired Comput Theor Appl BIC-TA 2010 2010:1510–7. https://doi.org/10.1109/BICTA.2010.5645272.
• [21] Maitra IK, Nag S, Bandyopadhyay SK. Technique for preprocessing of digital mammogram. Comput Methods Programs Biomed 2012;107:175–88. https://doi.org/10.1016/j.cmpb.2011.05.007.
• [22] Vikhe PS, Thool VR. Intensity based automatic boundary identification of pectoral muscle in mammograms. Procedia Comput Sci 2016;79:262–9. https://doi.org/10.1016/j.procs.2016.03.034.
• [23] Taifi K, Ahdid R, Fakir M, Elbalaoui A, Safi S, Taifi N. Automatic breast pectoral muscle segmentation on digital mammograms using morphological watersheds. Proc - 2017 14th Int Conf Comput Graph Imaging Vis CGiV 2017 2018; 1:126–31. https://doi.org/10.1109/CGiV.2017.24.
• [24] Singh VP, Srivastava R. Automated and effective content-based mammogram retrieval using wavelet based CS-LBP feature and self-organizing map. Biocybern Biomed Eng 2018;38:90–105. https://doi.org/10.1016/j.bbe.2017.09.003.
• [25] Rampun A, Morrow PJ, Scotney BW, Winder J. Fully automated breast boundary and pectoral muscle segmentation in mammograms. Artif Intell Med 2017;79:28–41. https://doi.org/10.1016/j.artmed.2017.06.001.
• [26] Devi SS, Vidivelli S. Classification of breast tissue density in digital mammograms. Proc 2017 Int Conf Innov Information, Embed Commun Syst ICIIECS 2017 2018; 2018-January: 1–7. https://doi.org/10.1109/ICIIECS.2017.8276139.
• [27] Liu Q, Liu L, Tan Y, Wang J, Ma X, Ni H. Mammogram density estimation using sub-region classification. Proc - 2011 4th Int Conf Biomed Eng Informatics, BMEI 2011 2011; 1:356–9. https://doi.org/10.1109/BMEI.2011.6098327.
• [28] Subashini TS, Ramalingam V, Palanivel S. Automated assessment of breast tissue density in digital mammograms. Comput Vis Image Underst 2010;114:33–43. https://doi.org/10.1016/j.cviu.2009.09.009.
• [29] Bora VB, Kothari AG, Keskar AG. Robust automatic pectoral muscle segmentation from mammograms using texture gradient and euclidean distance regression. J Digit Imaging 2016;29:115–25. https://doi.org/10.1007/s10278-015-9813-5.
• [30] Ferrari RJ, Rangayyan RM, Desautels JEL, Borges RA, Frere AF. Automatic identification of the pectoral muscle in mammograms. IEEE Trans Med Imaging 2004;23:232–45. https://doi.org/10.1109/TMI.2003.823062.
• [31] Mustra M, Bozek J, Grgic M. Breast border extraction and pectoral muscle detection using wavelet decomposition. Ieee Eurocon 2009, Eurocon 2009 2009:1426–33. https://doi.org/10.1109/EURCON.2009.5167827.
• [32] Kumar I, H.S. B, Virmani J, Thakur S. A classification framework for prediction of breast density using an ensemble of neural network classifiers. Biocybern. Biomed Eng 2017;37:217–28. https://doi.org/10.1016/j.bbe.2017.01.001.
• [33] Wang K, Khan N, Chan A, Dunne J, Highnam R. Deep learning for breast region and pectoral muscle segmentation in digital mammography. Lect Notes Comput Sci (Including Subser Lect Notes Artif Intell Lect Notes Bioinformatics) 2019; 11854 LNCS: 78–91. https://doi.org/10.1007/978-3-030-34879-3_7.
• [34] Ali MJ, Raza B, Shahid AR, Mahmood F, Yousuf MA, Dar AH, et al. Enhancing breast pectoral muscle segmentation performance by using skip connections in fully convolutional network. Int J Imaging Syst Techno 2020;30:1108–18. https://doi.org/10.1002/ima.22410.
• [35] Kim YJ, Yoo EY, Kim KG. Deep learning based pectoral muscle segmentation on Mammographic Image Analysis Society (MIAS) mammograms. Precis Futur Med 2021;5:77–82.
• [36] Guo Y, Zhao W, Li S, Zhang Y, Lu Y. Automatic segmentation of the pectoral muscle based on boundary identification and shape prediction. Phys Med Biol 2020;65:045016. https://doi.org/10.1088/1361-6560/ab652b.
• [37] M. Heath, K. Bowyer, D. Kopans RM and PKJ. The Digital Database for Screening Mammography. Fifth Int Work Digit Mammography, MJ Yaffe, Ed, Med Phys Publ 2001 2001:212–8.
• [38] Bt Ahmad SA, Taib MN, Khalid NEA, Taib H. Analysis of image quality based on dentists’ perception cognitive analysis and statistical measurements of intra-oral dental radiographs. 2012 Int Conf Biomed Eng ICoBE 2012 2012:379–84. https://doi.org/10.1109/ICoBE.2012.6179042.
• [39] Zeng M, Li Y, Meng Q, Yang T, Liu J. Improving histogram-based image contrast enhancement using gray-level information histogram with application to X-ray images. Optik (Stuttg) 2012;123:511–20. https://doi.org/10.1016/j.ijleo.2011.05.017.
• [40] Öktem H, Egiazarian K, Niittylahti J, Lemmetti J. An approach to adaptive enhancement of diagnostic X-ray images. EURASIP J Appl Signal Processing 2003;2003:430–6. https://doi.org/10.1155/S1110865703211069.
• [41] Deng G. A generalized unsharp masking algorithm. IEEE Trans Image Process 2011;20:1249–61. https://doi.org/10.1109/TIP.2010.2092441.
• [42] Chang DC, Wu WR. Image contrast enhancement based on a histogram transformation of local standard deviation. IEEE Trans Med Imaging 1998;17:518–31. https://doi.org/10.1109/42.730397.
• [43] Huang SC, Cheng FC, Chiu YS. Efficient contrast enhancement using adaptive gamma correction with weighting distribution. IEEE Trans Image Process 2013;22:1032–41. https://doi.org/10.1109/TIP.2012.2226047.
• [44] Slavkovic-Ilic M, Gavrovska A, Milivojevic M, Reljin I, Reljin B. Breast region segmentation and pectoral muscle removal in mammograms. Telfor J 2016;8:50–5. https://doi.org/10.5937/telfor1601050S.
• [45] Tarjan R. Depth-first search and linear graph algorithms Connected Compon. SIAM J Comput 1972;1:146–60.
• [46] Liu CC, Tsai CY, Liu J, Yu CY, Yu SS. A pectoral muscle segmentation algorithm for digital mammograms using Otsu thresholding and multiple regression analysis. Comput Math with Appl 2012;64:1100–7. https://doi.org/10.1016/j.camwa.2012.03.028.
• [47] Selvan S, Shenbaga Devi S. Automatic seed point selection in ultrasound echography images of breast using texture features. Biocybern Biomed Eng 2015;35:157–68. https://doi.org/10.1016/j.bbe.2014.10.001.
• [48] Raba D, Oliver A, Martí J, Peracaula M, Espunya J. Breast segmentation with pectoral muscle suppression on digital mammograms. Lect Notes Comput Sci 2005;3523:471–8. https://doi.org/10.1007/11492542_58.
• [49] Rampun A, López-Linares K, Morrow PJ, Scotney BW, Wang H, Ocaña IG, et al. Breast pectoral muscle segmentation in mammograms using a modified holistically-nested edge detection network. Med Image Anal 2019;57:1–17. https://doi. org/10.1016/j.media.2019.06.007.

Uwagi

Opracowanie rekordu ze środków MNiSW, umowa Nr 461252 w ramach programu "Społeczna odpowiedzialność nauki" - moduł: Popularyzacja nauki i promocja sportu (2021).