Automatic Detection of Microcalcifications in ROI Images Based on PFCM and ANN

Abstract

This paper presents a novel method for the automatic detection of microcalcifications in regions of interest images. Automatic detection method is implemented by feature extraction and sub-segmentation steps. The feature extraction step is improved using a top-hat transform such that microcalcifications can be highlighted. In a second step a sub-segmentation method based on the possibilistic fuzzy c-means clustering algorithm is applied in order to segment the images and as a way to identify the atypical pixels inside the regions of interest as the pixels representing microcalcifications. Once the pixels representing these objects have been identified, an artificial neural network model is used to learn the relations between atypical pixels and microcalcifications, such that the model can be used for aid diagnosis, and a medical could determine if these regions of interest are benign or malignant. So, as the results show, the proposed approach is a good alternative for the detection of suspicious regions, which could be of great help for medical diagnosis.

Keywords::

• Microcalcifications Detection
• Possibilistic Fuzzy C-Means
• Artificial Neural Network

Additional information

Notes on contributors

J. Quintanilla-Domínguez

J. Quintanilla-Domínguez is a Full Professor of Technological University of the Southwest of Guanajuato, Mexico. He received the degree of BSc in Electronics and Telecommunications Engineering and MSc in Electrical Engineering from University of Guanajuato in 2004 and 2007, respectively. He is studying his PhD at the Technical University of Madrid (UPM). He received the price for best paper in the 7th IEEE international Conference on Industrial Informatics, INDIN 2009. His research interests include Pattern Recognition and Classification, Image Processing, Artificial Neural Networks and Fuzzy Set Techniques.

B. Ojeda-Magaña

B. Ojeda-Magaña was born in 1976 in Guadalajara, Mexico. He received his BSc in Civil Engineering and MsC in Projects Engineering from the University of Guadalajara, Mexico, in 2003 and 2006, respectively. He achieved the PhD degree in 2010 in Telecommunication Engineer by the Technical University of Madrid (UPM). He received the price for best paper in the 7th IEEE international Conference on Industrial Informatics, INDIN 2009. His research interests are Pattern Recognition, based mainly on Fuzzy Set Techniques, as well as Artificial Neural Networks, Pattern Classification, Data Mining and Image Processing.

A. Marcano-Cedeño

A. Marcano-Cedeño was born in Bolivar city, Venezuela. He received the degree of BSc in Computer Science by Central University of Venezuela (UCV), in 1995. He received the Master degree information Systems in Andres Bello Catholic University in 1998. In 2000 he is an assistant professor in the Pedagogical University Experimental Libertador (UPEL). In 2005 He achieved Master degree in Communication Systems and Networks by the Technical University of Madrid (UPM). He achieved the PhD degree in 2010 in Telecommunication Engineer by the Technical University of Madrid (UPM). He presently works for Biomedical Engineering and Telemedicine Centre of UPM (GBT/UPM). His research interests include Artificial Neural Networks, Pattern Classification, Data Mining, Image Processing and Analysis, Bioinformatics System and Telemedicine.

J.M. Barrón-Adame

J.M. Barrón-Adame is a Full Professor of Technological University of the Southwest of Guanajuato, Mexico. He was born in 1972 in Guanajuato, Mexico. He received his BS degree in 1998 by the Faculty of Mechanical, Electrical and Electronic Engineering (FIMEE). In 2008 he won a scholarship awarded by the National Council on Science and Technology (CONACyT) in Telecommunication Engineer from the Technical University of Madrid (UPM) where he received the PhD degree in 2010. His research interests are Pattern Recognition, Pattern Classification, Data Mining and Image Processing based mainly on Artificial Neural Networks.

A. Vega-Corona

A. Vega-Corona. He received the degree of BSc in Electronics and Telecommunications Engineering and MSc in Electrical Engineering from University of Guanajuato in 1989 and 1993, respectively. He achieved the PhD degree in 2004 in Telecommunication Engineer by the Technical University of Madrid (UPM). His research interests are Pattern Recognition, Signals Processing and Artificial Neural Networks.

D. Andina

D. Andina (Prof Dr-Eng, IEEE Senior Member), was born in Madrid, Spain, where he received simultaneously two Master degrees, on Computer Science and on Electronics and Communications by the Technical University of Madrid (UPM), Spain, in 1990. He achieved the PhD degree in 1995 with a thesis on Artificial Neural Networks applications in Signal Processing. He presently works for UPM where he heads the Group for Automation in Signals and Communications (GASC/ UPM), are search group interested in Signal Processing and Computational Intelligence Applications: Man—Machine Systems and Cybernetics. He is author or co-author of more than 200 national and international publications, being director of more than 50R+D projects financed by National and Local Governments, European Commission or private Institutions and Firms. He is also Associate Editorial Member of several International Research Journals and Transactions, and has participated in the organization of more than 50 international Research, Innovation or Technology Transfer events.