Tolga YÜKSEL, Semih ERGİN, Idil ISIKLI ESENER

A novel multistage system for the detection and removal of pectoral muscles in mammograms

In this paper, a novel multistage scheme for pectoral muscle removal from mammography images is proposed, and the performance of this system is veri ed using the publicly available Mammographic Image Analysis Society digital mammogram database. This database is composed of mediolateral oblique mammography images including three different tissue types (fatty, fatty-glandular, and dense-glandular) with three health status types (normal, benign cancer, and malignant cancer). In the implementation of the proposed system, a mammography image is rst preprocessed by performing noise reduction background removal followed by artifact suppression processes. Then a presegmentation procedure is applied using region growing and line tting is executed. Finally, pixels including pectoral muscle regions are removed from mammography images with an accuracy of 94.40%, sensitivity of 89.62%, and speci city of 99.99% after some postprocessing operations. Although the mean false positive rate obtained by the proposed approach is higher than that of other studies in the literature, not only the lower mean false negative rate but also the enhancement in the quality of pectoral muscle segmentation for all 322 images in the database evidently show the success of the proposed approach.

PDF

___

[1] Ganesan K, Acharya UR, Chua KC, Min LC, Abraham KT. Pectoral muscle segmentation: A review. Comput Meth Prog Bio 2013; 110: 48-57.
[2] Saltanat M, Hossain MA, Alam MS. An efficient pixel value based mapping scheme to delineate pectoral muscle from mammograms. In: IEEE 2010 International Conference on Bio-Inspired Computing: Theories and Applications; 23{26 September 2010; Changsha, China. New York, NY, USA: IEEE. pp. 1510-1517.
[3] Roshan DY, Harada K. A connected component labelling algorithm for grayscale images and application of the Algorithm on mammograms. In: 2007 Annual ACM Symposium on Applied Computing; 11{15 March 2007; Seoul, Republic of Korea. New York, NY, USA: ACM. pp. 146-152.
[4] Nagi J, Kareem SA, Nagi F, Ahmed SK. Automated breast pro le segmentation for ROI detection using digital mammograms. In: IEEE 2010 EMBS Conference on Biomedical Engineering & Sciences; 30 November{2 December 2010; Kuala Lumpur, Malaysia. New York, NY, USA: IEEE. pp. 87-92.
[5] Liu L, Wang J, He K. Breast density classi cation using histogram moments of multiple resolution histograms. In: 2010 International Conference on Biomedical Engineering and Informatics; 16{18 October 2010; Yantai, China. New York, NY, USA: IEEE. pp. 146-149.
[6] 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 Appl 2012; 64: 1100-1107.
[7] Kurt B, Nabiyev VV, Turhan K. A novel automatic suspicious mass regions identi cation using Havrda & Charvat entropy and Otsu's N thresholding. Comput Meth Prog Bio 2014; 114: 349-360.
[8] Thangavel K, Karnan M. Computer aided diagnosis in [15] Ma WY, Manjunath BS. Edge ow: A technique for boundary detection and image segmentation. IEEE T Image Process 2000; 9: 1375-1388.
[8] Thangavel K, Karnan M. Computer aided diagnosis in digital mammograms: Detection of microcalci cations by meta heuristic algorithms. Int J Artif Intell Mach Learn 2005; 5: 29-40.
[9] David R, Arnau O, Joan M, Marta P, Joan E. Breast segmentation with pectoral muscle suppression on digital mammograms. Lect Notes Comput Sc 2005; 153-158.
[10] Subashini T S, Ramalingam V, Palanivel S. Pectoral muscle removal and detection of masses in digital mammogram using CCL. Int J Comput Appl 2010; 1: 71-76.
[11] Chakraborty J, Mukhopadhyay S. Automatic detection of pectoral muscle using average gradient and shape based feature. J Digit Imaging 2012; 25: 387-399.
[12] Camilus KS, Govindan VK, Sathidevi PS. Pectoral muscle identi cation in mammograms. J Appl Clin Med Phys 2011; 12: 215-230.
[13] Mustra M, Bozek J, Grgic M. Breast border extraction and pectoral muscle detection using wavelet decomposition. In: IEEE 2009 International Conference on Computer as a Tool; 18{23 May 2009; St. Petersburg, Russia. New York, NY, USA: IEEE. pp. 1426-1433.
[14] Ferrari RJ, Rangayyan RM, Desautels JEL, Borges RA, Frere AF. Automatic identi cation of the pectoral muscle in mammograms. IEEE T Med Imaging 2004; 23: 232-245.
[16] Karssemeijer N. Automated classi cation of parenchymal patterns in mammograms. Phys Med Biol 1998; 43: 365- 378.
[17] Kwok SM, Chandrasekhar R, Attikiouzel Y. Automatic pectoral muscle segmentation on mammograms by straight line estimation and cliff detection. In: IEEE 2001 Australian and New Zealand Intelligent Information Systems Conference; 18{21 November 2001; Perth, Western Australia. New York, NY, USA: IEEE. pp. 67-72.
[18] Kwok SM, Chandrasekhar R, Attikiouzel Y, Rickard MT. Automatic pectoral muscle segmentation on mediolateral oblique view mammograms. IEEE T Med Imaging 2004; 23: 1129-1140.
[19] Mustra M, Grgic M. Robust automatic breast and pectoral muscle segmentation from scanned mammograms. Signal Process 2013; 93: 2817-2827.
[20] Tzikopoulos SD, Mavroforakis ME, Georgiou HV, Dimitropoulos N, Theodoridis S. A fully automated scheme for mammographic segmentation and classi cation based on breast density and asymmetry. Comput Meth Prog Bio 2011; 102: 47-63.
[21] Vaidehi K, Subashini TS. Automatic identi cation and elimination of pectoral muscle in digital mammograms. Int J Comput Appl 2013; 75: 15-18.
[22] Yam M, Brady M, Highnam R, Behrenbruch C, English R, Kita Y. Three-dimensional reconstruction of microcal- ci cation clusters from two mammographic views. IEEE T Med Imaging 2001; 20: 479-489.
[23] Weidong X, Lihua L, Wei LA. Novel pectoral muscle segmentation algorithm based on polyline tting and elastic thread approaching. In: 2007 International Conference on Bioinformatics and Biomedical Engineering; 6{8 July 2007; Wuhan, China. New York, NY, USA: IEEE. pp. 837-840.
[24] Domingues I, Cardoso JS, Amaral I, Moreira I, Passarinho P, Comba JS, Correia R, Cardoso MJ. Pectoral muscle detection in mammograms based on the shortest path with endpoints learnt by SVMs. Eng Med Biol 2010; 3158- 3161.
[25] Camilus KS, Govindan VK, Sathidevi PS. Computer-aided identi cation of the pectoral muscle in digitized mam- mograms. J Digit Imaging 2010; 23: 562-580.
[26] Abdellatif H, Taha TE, Zahran OF, Al-Nauimy W, El-Samie FEA. Automatic pectoral muscle boundary detection in mammograms using eigenvectors segmentation. In: 2012 International Radio Science Conference; 10{12 April 2012; Cairo, Egypt. New York, NY, USA: IEEE. pp. 633-640.
[27] Liu L, Wang J, Wang T. Breast and pectoral muscle contours detection based on goodness of t measure. In: 2011 International Conference on Bioinformatics and Biomedical Engineering; 10{12 May 2011; Wuhan, China. New York, NY, USA: IEEE. pp. 1-4.
[28] Moayedi F, Azimifar Z, Boostani R, Katebi S. Contourlet-based mammography mass classi cation using the SVM family. Comput Biol Med 2010; 40: 373-383.
[29] Sultana A, Ciuc M, Strungaru R. Detection of pectoral muscle in mammograms using mean-shift segmentation approach. In: IEEE 2010 International Conference on Communications; 23{27 May 2010; Cape Town, South Africa. New York, NY, USA: IEEE. pp. 165-168.
[30] Li JB. Mammographic image based breast tissue classi cation with kernel self-optimized sher discriminant for breast cancer diagnosis. J Med Syst 2012; 36: 2235-2244.
[31] Maitra IK, Nag S, Bandyopadhyay SK. Technique for preprocessing digital mammogram. Comput Meth Prog Bio 2012; 107: 175-188.
[32] Talha M, Sulong GB. Preprocessing and pectoral muscle separation from breast mammograms. Int J Phys Sci 2012; 7: 471-477.
[33] Anitha J, Peter JD, Pandian SIA. A dual stage adaptive thresholding (DuSAT) for automatic mass detection in mammograms. Comput Meth Prog Bio 2017; 138: 93-104.digital mammograms: Detection of microcalci cations by meta heuristic algorithms. Int J Artif Intell Mach Learn 2005; 5: 29-40.
[34] Guo Y, Dong M, Yang Z, Gao X, Wang K, Luo C, Ma Y, Zhang J. A new method of detecting micro-calci cation clusters in mammograms using contourlet transform and non-linking simpli ed PCNN. Comput Meth Prog Bio 2016; 130: 31-45.
[35] Taghanaki SA, Liu Y, Miles B, Hamarneh G. Geometry based pectoral muscle segmentation from MLO mammogram views. IEEE T Bio-Med Eng 2017; in press.
[36] Feudjio CK, Klein J, Tiedeu A, Colt O. Automatic extraction of pectoral muscle in the MLO view of mammograms. Phys Med Biol 2013; 58: 8493.
[37] Gonzalez RC, Woods RE. Image restoration. In: Digital Image Processing. 2nd ed. Upper Saddle River, NJ, USA: Prentice Hall, 2004, pp. 241-243.
[38] Otsu N. A threshold selection method from gray-level histograms. IEEE T Syst Man Cyb 1979; 9: 62-66.
[39] Gonzalez RC, Woods RE. Image segmentation. In: Digital Image Processing. 3rd ed. Upper Saddle River, NJ, USA: Prentice Hall, 2008. pp. 711-816.
[40] Suckling J, Parker J, Dance DR, Astley S, Hutt I, Boggis CRM, Ricketts I, Stamatakis E, Cerneaz N, Kok SL et al. The mammographic image analysis society digital mammogram database. Exerpta Medica 1994; International Congress Series 1069: 375-378.
[41] Ma F, Bajger M, Slavotinek JP, Bottema MJ. Two graph theory based methods for identifying the pectoral muscle in mammograms. Pattern Recogn 2007; 40: 2592-2602.