روش جديد آسيب‌شناسي توده‌ها در تصاوير ماموگرافي به كمک ترکیب ويژگي‌هاي منطبق بر استاندارد BI - RADS و كلاسه‌بندي كننده مبتني بر تضاد
محورهای موضوعی : مهندسی برق و کامپیوتر
فاطمه ساکی
1
(دانشگاه علم و صنعت ايران)
امیر طهماسبی
2
(دانشگاه علم و صنعت ايران)
شهریار برادران شکوهی
3
(مهندسی برق)
کلید واژه: استخراج ويژگي استاندارد BI-RADS سيستم CADx کلاسه‌بندی کننده مبتنی بر تضاد ماموگرافي,
چکیده مقاله :
تفكيك توده‌هاي خوش‌خيم و بدخيم در ماموگرام‌هاي ديجيتالي يكي از مراحل بسيار مهم تشخيص زودهنگام سرطان سينه است، چرا كه مي‌تواند تا حد زيادي شانس بقاي بيمار را افزايش دهد. در اين مقاله يك سيستم CADx نوين با به‌کارگيري کلاسه‌بندي کننده جديد مبتني بر تضاد (OWBP) جهت آسيب‌شناسي توده‌ها در تصاوير ماموگرافي معرفي خواهد شد. هدف، بهبود عملکرد و سرعت يادگيري الگوريتم‌هاي CADx با استفاده از ترکيب ويژگي‌هاي منطبق بر استاندارد BI-RADS و كلاسه‌بندي كننده پيشنهادي مي‌باشد. ورودي سيستم يک ROI بوده که حاوي يک توده مشکوک است. اين ناحيه ابتدا تحت پيش‌پردازش‌هايي قرار گرفته، سپس 12 ويژگي که توصيف‌کننده‌هاي مناسبي از شکل، مرز و چگالي توده هستند، استخراج مي‌شوند. منحنی ROC و عملكرد آسيب‌شناسي حاصل از ترکيب تمام اين ويژگي‌ها توسط دو کلاسه‌بندي کننده با يادگيري متداول پس‌انتشار و يادگيري پيشنهادي OWBP ارزيابي شده و سيستمهاي حاصل از لحاظ سرعت يادگيري نیز مورد مقايسه قرار گرفته‌اند. همچنین در اين تحقيق قابليت آسيب‌شناسي هر گروه از ويژگي‌هاي شكل، مرز و چگالي به‌طور جداگانه بررسي شده است. پايگاه داده مورد استفاده در اين تحقيق MIAS است. سيستم نهايي پیشنهادی داراي Az 924/0، با سرعت يادگيري تقريباً 4 برابر سرعت يادگيري سيستم با کلاسه‌بندي کننده پس‌انتشار و همچنين عملکرد 86/92% مي‌باشد.
Fast and accurate classification of benign and malignant patterns in digital mammograms is of significant importance in the diagnosis of breast cancers. In this paper, we develop a new Computer-aided Diagnosis (CADx) system using a novel Opposition-based classifier to enhance the accuracy and shorten the training time of the classification of breast masses. We extract a group of Breast Imaging-Reporting and Data System (BI-RADS) features from preprocessed mammography images and feed them to a Multi-Layer Perceptron (MLP). The MLP is then trained using a new learning rule which we will refer to as the Opposite Weighted Back Propagation (OWBP) algorithm. We evaluate the performance of the system, in terms of classification accuracy, using a Receiver Operational Characteristics (ROC) curve. The proposed system yields an area under ROC curve (Az) of 0.924 and an accuracy of 92.86 %. Furthermore, the speed analysis results suggest that, with the same network topology, the convergence rate of the proposed OWBP algorithm is almost 4 times faster than that of the traditional Back Propagation (BP) algorithm.
[1] American Cancer Society, Breast Cancer Facts & Figures 2009-2010, Atlanta, 2009.
[2] A. C. Bovik, Handbook of Image and Video Processing, 2nd ed., Elsevier Academic Press, pp. 1195-1217, 2005.
[3] R. M. Rangayyana, F. J. Ayresa, and J. E. L. Desautels, "A review of computer aided diagnosis of breast cancer: toward the detection of subtle signs," J. of the Franklin Institute, vol. 344, no. 3-4, pp. 312-348, May/Jul. 2007.
[4] H. D. Cheng, X. J. Shi, R. Min, L. M. Hu, X. P. Cai, and H. N. Du, "Approaches for automated detection and classification of masses in mammograms," J. of Pattern Recognition, vol. 39, no. 4, pp. 646-668, 2006.
[5] American College of Radiology, ACR BI-RADS-Mammography, Ultrasound & Magnetic Resonance Imaging, 4th ed., American College of Radiology, Reston, VA, 2003.
[6] H. R. Tizhoosh, "Opposition-based learning: a new scheme for machine intelligence," in Proc. of the IEEE, International Conf. on Computational Intelligence for Modeling, Control, and Automation, vol. 1, pp. 695-701, Vienna, Austria, 2005.
[7] M. Ventreca and H. R. Tizhoosh, "Improving the convergence of back propagation by opposite transfer function," in Proc. of Int. Joint Conf. on Neural Networks, pp. 4777-4784, 2006.
[8] H. R. Tizhoosh and M. Ventresca, Oppositional Concepts in Computational Intelligence, Springer, 2008.
[9] F. Saki, A. Tahmasbi, and S. B. Shokouhi, "A novel opposition-based classifier for mass diagnosis in mammography images," in Proc. of the IEEE, 17th Iranian Conf. on Biomedical Eng. ICBME'2010, 4 pp., Isfahan, Iran, 3-4 Nov. 2010.
[10] X. Zhang, X. Gao, and Y. Wang, "MCs detection with combined image features and twin support vector machines," J. of Computers, vol. 4, no. 3, pp. 215-221, Mar. 2009.
[11] A. Tahmasbi, F. Saki, and S. B. Shokouhi, "Classification of benign and malignant masses based on zernike moments," J. of Computers in Biology and Medicine, vol. 41, no. 8, pp. 726-735, 2010.
[12] J. Suckling et al., "The mammographic image analysis society digital mammogram database," Exerpta Medical, Int. Congress Series 1069, pp. 375-378, 1994.
[13] L. M. Bruce and R. R. Adhami, "Classifying mammographic mass shapes using the wavelet transform modulus-maxima method," IEEE Trans. on Medical Imaging, vol. 18, no. 12, pp. 1170-1177, Dec. 1999.
[14] W. Wang, J. E. Mottershead, and C. Mares, "Mode-shape recognition and finite element model updating using the zernike moment descriptor," J. of Mechanical Systems and Signal Processing, vol. 23, no. 7, pp. 2088-2112, Oct. 2009.
[15] W. K. Pratt, Digital Image Processing: PIKS Inside, 3rd Edition, John Wiley & Sons, Ch. 18, 2001.
[16] C. Balleyguier et al., "BIRADSTM classification in mammography," European J. of Radiology, vol. 61, no. 2, pp. 192-194, Feb. 2007.
[17] A. Tahmasbi, F. Saki, and S. B. Shokouhi, "An effective breast mass diagnosis system using Zernike moments," in Proc. 17th Iranian Conf. on Biomedical Engineering, ICBME'2010, 4 pp., Isfahan, Iran, 2010.
[18] A. Tahmasbi, F. Saki, H. Aghapanah, and S. B. Shokouhi, " A novel breast mass diagnosis system based on Zernike moments as shape and density descriptors," in Proc. 18th Iranian Conf. on Biomedical Engineering, ICBME'2011, pp. 100-104, Tehran, Iran, 2011.
[19] K. V. Augustine and H. Dongjun, "Image similarity for rotation invariants image retrieval system," in Proc. of the IEEE, Int. Conf. on Multimedia Computing and System, ICMCS'09, pp. 133-137, 2-4 Apr. 2009.
[20] B. Verma, P. McLeod, and A. Klevansky, "Classification of benign and malignant patterns in digital mammograms for the diagnosis of breast cancer," J. of Expert Systems with Application, vol. 37, no. 4, pp. 3344-3351, Apr. 2010.
[21] F. O. Karray and C. de Silva, Soft Computing and Intelligent Systems Design, Pearson Education Limited, 2004.
[22] J. Bozek, M. Mustra, K. Delac, and M. Grgic, "A survey of image processing algorithms in digital mammography," in Recent Advances in Multimedia Signal Processing and Communication, vol. 231, pp. 631-657, Oct. 2009.
[23] A. Tahmasbi, F. Saki, and S. B. Shokouhi, "Mass diagnosis in mammography images using novel FTRD features," in Proc. of the IEEE, 17th Iranian Conf. on Biomedical Engineering, ICBME'2010, Isfahan, Iran, 5 pp., 3-4 Nov. 2010.
[24] A. M. Abdalla et al., "Breast cancer detection based on statistical textural features classification," in Proc. of the IEEE, 4th Int. Conf. on Innovations in Information Technology, pp. 728-730, 18-20 Nov. 2007.
[25] D. Manrique, J. Rios, and A. Rodriguez - Paton, "Evolutionary system for automatically constructing and adapting radial basis function networks," J. of Neurocomputing, vol. 69, no. 4, pp. 2268-2283, Apr. 2006.
[26] N. R. Mudigonda, R. M. Rangayyan, and J. E. L. Desautels, "Gradient and texture analysis for the classification of mammographic masses," IEEE Trans. on Medical Imaging, vol. 19, no. 10, pp. 1032-1043, Oct. 2000.