Vitoantonio Bevilacqua
ABSTRACT
Introduction and objective: Computer Aided Decision (CAD) systems based on Medical Imaging could support radiologists in classifying malignant regions from benign ones, in the field of investigation for breast cancer detection. This decision may often follow a previous procedure dedicated to the earlier identification of Regions Of Interest (ROI) containing still unclassified lesions.
Materials and methods: Materials comprise features extracted from magnetic resonance (MR) images representing morphological properties of lesions. The Regions Of Interest identified by a previous automatic procedure validated by radiologists of the University of Bari Aldo Moro (Italy), authors of this work, 134 from 600 slices considered of interest, because they contain still unclassified damaged areas. Several techniques were tested for ROI segmentation and classification. In particular, it can be shown that the same procedures for lesioned-area discrimination were also useful for malignancy classification of lesions, themselves. In particular, MR images were processed with different image processing techniques for ROI extraction, which were, ultimately, described by morphological features, such as circularity, aspect ratio, solidity and convexity. Finally, we discuss a procedure to design a feed-forward supervised artificial neural networks (ANN) architecture based on an evolutionary strategy. In a similar approach, different ANN topologies were tested in order to find the best in terms of mean accuracy for several iterations of training, validation and test. In particular, for each topology, the training, validation and test sets were constructed using 100 random permutations of the dataset, from which the average performances were calculated.
Results: The performance of the best ANN architecture, trained using a training set of 82 samples (equally divided between malignant and benign lesions) from the 134 samples available in the whole dataset, were evaluated in terms of accuracy, sensitivity and specificity.
Conclusion: Testing determined that the supervised ANN approach is consistent and reveals good performance; in particular, the optimal ANN topology found through an evolutionary strategy showed high generalization on the mean performance indexes regardless of training, validation and test sets applied, showing good performances in terms of both accuracy and sensitivity, permitting correct classification of the true malignant lesions.
- US Cancer Statistics Working Group et al. United states cancer statistics: 1999--2010 incidence and mortality web-based report. Atlanta: US Department of Health and Human Services, Centers for Disease Control and Prevention and National Cancer Institute, 2013.Google Scholar
- William H Wolberg, W Nick Street, and Olvi L Mangasarian. Image analysis and machine learning applied to breast cancer diagnosis and prognosis. Analytical and Quantitative cytology and histology, 17(2):77--87, 1995.Google Scholar
- William H Wolberg, W Nick Street, and Olvi L Mangasarian. Breast cytology diagnosis via digital image analysis. Analytical and Quantitative Cytology and Histology, 15(6):396--404, 1993.Google Scholar
- Vitoantonio Bevilacqua. Three-dimensional virtual colonoscopy for automatic polyps detection by artificial neural network approach: New tests on an enlarged cohort of polyps. Neurocomputing, 116:62--75, 2013.Google ScholarCross Ref
- Olvi L Mangasarian, W Nick Street, and William H Wolberg. Breast cancer diagnosis and prognosis via linear programming. Operations Research, 43(4):570--577, 1995. Google ScholarDigital Library
- Vitoantonio Bevilacqua, Giuseppe Mastronardi, and Filippo Menolascina. Hybrid data analysis methods and artificial neural network design in breast cancer diagnosis: Idest experience. In Computational Intelligence for Modelling, Control and Automation, 2005 and International Conference on Intelligent Agents, Web Technologies and Internet Commerce, International Conference on, volume 2, pages 373--378. IEEE, 2005. Google ScholarDigital Library
- William H Wolberg, W Nick Street, Dennis M Heisey, and Olvi L Mangasarian. Computer-derived nuclear features distinguish malignant from benign breast cytology. Human Pathology, 26(7):792--796, 1995.Google ScholarCross Ref
- Vitoantonio Bevilacqua, Giuseppe Mastronardi, Filippo Menolascina, Paolo Pannarale, and Antonio Pedone. A novel multi-objective genetic algorithm approach to artificial neural network topology optimisation: the breast cancer classification problem. In Neural Networks, 2006. IJCNN'06. International Joint Conference on, pages 1958--1965. IEEE, 2006.Google Scholar
- Sylvia H Heywang-Köbrunner and Richard W Katzberg. Contrast-enhanced magnetic resonance imaging of the breast. Investigative radiology, 29(1):94--104, 1994.Google ScholarCross Ref
- Susan G Orel, Mitchell D Schnall, Virginia A LiVolsi, and Rosalind H Troupin. Suspicious breast lesions: Mr imaging with radiologic-pathologic correlation. Radiology, 190(2):485--493, 1994.Google ScholarCross Ref
- Wei Huang, Paul R Fisher, Khaldoon Dulaimy, Luminita A Tudorica, Brian O'Hea, and Terry M Button. Detection of breast malignancy: Diagnostic mr protocol for improved specificity 1. Radiology, 232(2):585--591, 2004.Google ScholarCross Ref
- CK Kuhl. Mri of breast tumors. European radiology, 10(1):46--58, 2000.Google ScholarCross Ref
- F Baum, U Fischer, R Vosshenrich, and E Grabbe. Classification of hypervascularized lesions in ce mr imaging of the breast. European radiology, 12(5):1087--1092, 2002.Google ScholarCross Ref
- Ulf Hoffmann, Gunnar Brix, Michael V Knopp, Thomas Heβ, and Walter J Lorenz. Pharmacokinetic mapping of the breast: a new method for dynamic mr mammography. Magnetic resonance in medicine, 33(4):506--514, 1995.Google Scholar
- Hadassa Degani, Vadim Gusis, Daphna Weinstein, Scott Fields, and Shalom Strano. Mapping pathophysiological features of breast tumors by mri at high spatial resolution. Nature medicine, 3(7):780--782, 1997.Google ScholarCross Ref
- Susan G Orel. High-resolution mr imaging for the detection, diagnosis, and staging of breast cancer. Radiographics, 18(4):903--912, 1998.Google ScholarCross Ref
- Paolo Belli, Melania Costantini, Enida Bufi, Andrea Magistrelli, Giuseppe La Torre, and Lorenzo Bonomo. Diffusion-weighted imaging in breast lesion evaluation. La radiologia medica, 115(1):51--69, 2010.Google Scholar
- Sibel Kul, Aysegul Cansu, Etem Alhan, Hasan Dinc, Gurbuz Gunes, and Abdulkadir Reis. Contribution of diffusion-weighted imaging to dynamic contrast-enhanced mri in the characterization of breast tumors. American Journal of Roentgenology, 196(1):210--217, 2011.Google ScholarCross Ref
- Marco Moschetta, Michele Telegrafo, Leonarda Rella, Arcangela Capolongo, Amato Antonio Stabile Ianora, and Giuseppe Angelelli. Mr evaluation of breast lesions obtained by diffusion-weighted imaging with background body signal suppression (dwibs) and correlations with histological findings. Magnetic resonance imaging, 32(6):605--609, 2014.Google Scholar
- Edward S Fobben, Carole Z Rubin, Lester Kalisher, Alan G Dembner, Murray H Seltzer, and Elissa J Santoro. Breast mr imaging with commercially available techniques: radiologic-pathologic correlation. Radiology, 196(1):143--152, 1995.Google ScholarCross Ref
- Christiane Katharina Kuhl, Peter Mielcareck, Sven Klaschik, Claudia Leutner, Eva Wardelmann, Jürgen Gieseke, and Hans H Schild. Dynamic breast mr imaging: Are signal intensity time course data useful for differential diagnosis of enhancing lesions? 1. Radiology, 211(1):101--110, 1999.Google ScholarCross Ref
- Christiane K Kuhl, Hans H Schild, and Nuschin Morakkabati. Dynamic bilateral contrast-enhanced mr imaging of the breast: Trade-off between spatial and temporal resolution 1. Radiology, 236(3):789--800, 2005.Google ScholarCross Ref
- Judith MS Prewitt. Object enhancement and extraction. Picture processing and Psychopictorics, 10(1):15--19, 1970.Google Scholar
- David B Fogel. An information criterion for optimal neural network selection. IEEE transactions on neural networks/a publication of the IEEE Neural Networks Council, 2(5):490--497, 1990. Google ScholarDigital Library
- Filippo Menolascina, Domenico Bellomo, Thomas Maiwald, Vitoantonio Bevilacqua, Caterina Ciminelli, Angelo Paradiso, and Stefania Tommasi. Developing optimal input design strategies in cancer systems biology with applications to microfluidic device engineering. BMC Bioinformatics, 10(S-12):4, 2009.Google Scholar
- Vitoantonio Bevilacqua, Giuseppe Mastronardi, and Giuseppe Piscopo. Evolutionary approach to inverse planning in coplanar radiotherapy. Image Vision Comput., 25(2):196--203, 2007. Google ScholarDigital Library
- Martin Riedmiller and Heinrich Braun. A direct adaptive method for faster backpropagation learning: The rprop algorithm. In Neural Networks, 1993., IEEE International Conference on, pages 586--591. IEEE, 1993.Google ScholarCross Ref
- Kevin S Woods, Christopher C Doss, Kevin W Bowyer, Jeffrey L Solka, Carey E Priebe, and W Philip Kegelmeyer JR. Comparative evaluation of pattern recognition techniques for detection of microcalcifications in mammography. International Journal of Pattern Recognition and Artificial Intelligence, 7(06):1417--1436, 1993.Google ScholarCross Ref
- Nitesh V. Chawla, Kevin W. Bowyer, Lawrence O. Hall, and W. Philip Kegelmeyer. Smote: synthetic minority over-sampling technique. Journal of artificial intelligence research, pages 321--357, 2002. Google ScholarDigital Library
- Dennis Zill, Warren S Wright, and Michael R Cullen. Advanced engineering mathematics. Jones & Bartlett Learning, 2011. Google ScholarDigital Library
- Andrew P Bradley. The use of the area under the roc curve in the evaluation of machine learning algorithms. Pattern recognition, 30(7):1145--1159, 1997. Google ScholarDigital Library
Index Terms
- An Optimized Feed-forward Artificial Neural Network Topology to Support Radiologists in Breast Lesions Classification
Recommendations
Breast cancer classification using hybrid deep neural networks: staging and grading of cancer
Breast cancer is responsible for the deaths of hundreds of women every year. The manual identification of breast cancer has more difficulties, and have the possibility of error. Many imaging approaches are being researched for their potential to identify ...
Computer Aided Diagnosis System to Detect Breast Cancer Pathological Lesions
CIARP '08: Proceedings of the 13th Iberoamerican congress on Pattern Recognition: Progress in Pattern Recognition, Image Analysis and ApplicationsBreast cancer is one of the most frequent forms of women's cancer over the world. Studies of the World Health Organization (WHO) reported 1,151,298 cases in 2002. A reliable Computer-Aided-Diagnosis (CAD) system for automated detection/classification ...
An Efficient Feed Foreword Network Model with Sine Cosine Algorithm for Breast Cancer Classification
This article describes how breast cancer is the most common invasive cancer in females worldwide and is major cause of deaths. The diagnoses of breast cancer include mammograms, breast ultrasound, magnetic resonance imaging MRI, ductogram and biopsy. ...
Comments