Mammogram Detector: An Automated Masses Detecting Based on Case Based Reasoning Classifier and Harralix Texture Extrator for MRI Digital Breast Mammogram Diagnosis

The most important stage in reading the breast mammogram is to detect feature called masses. Masses are quite subtle to be seen on mammogram, and often occurred in the dense areas of the breast tissue, have smoother boundaries. It has many shapes such as circumscribed, speculated (or stellate), lobulated or ill-defined. Masses can be obscured or similar to normal breast parenchyma and yet is the most difficult to detect even by radiologist. Radiologists need training and experience on subjective criteria to be successfully detecting these masses on a mammogram. 

The objectives of this research are;

1. Digital mammogram masses detection algorithm.

The algorithm comprises of digital mammograms image enhancement and segmentation. Image enhancement is a process such as digitization and preprocessing to eliminate noise for better visual quality. Segmentation is a process to identify accurately each of distinct masses entity. There are 12-18 distinct mass features (global and local features) are extracted from the segmented masses.

    2. Designing a Case based reasoning (CBR) classifier approach for masses detection algorithm.

CBR is a way of solving a problem using information on past experiences and knowledge of old   previously solved problem. Case base classifier will be proposed which retrieves the most similar case to solve a new case. Similarity between two cases is computed using different similarity measures. The result obtained of this similarity measurement is evaluated by comparing the result of the classification with the diagnosis given by the biopsies.

 

Publication

Raman, V., Sumari, P., (2014), “MammoAPPS: Digital Mammogram Mobile Application for Diagnosing Breast Cancer and Retrieval Report for Decision Making to Medical Experts", in Journal of Advance Research in Computer Science and Software Engineering,Vol. 4 Issue 2.

Valliappan, R., Sumari, P., Then, P., Omari, S., (2011), “Review on Mammogram Mass Detection by Machine Learning Techniques”, International Journal of Computer and Electrical Engineering (IJCEE), ISSN: 1793-8198, Issue 5

Valliappan, R., Sumari, P., Then, T., (2011), ”Matlab Implementation and Results of Region Growing Segmentation Using Haralic Texture Features on Mammogram Mass Segmentation”, in Advances in Wireless, Mobile Networks and Applications, WiMoA 2011, pp 293- 303

Valliappan R., Patrick T., Sumari, P., (2011), “Mammogram Problem Solving Approach: Building CBR Classifier for Classification of Masses in Mammogram” in V.V. Das, J. Stephen, and Y, Chaba (Eds): CNC, CCIS 142, pp 284 – 289

Valliappan, R., Sumari, P., Leka, J. R., Dharamaprakash, G,. (2010), ”Performance Based CBR Mass Detection In Digital Mammograms” in IEEE ICCCT10, ISSN: 978-1-4244-7768-5, Tamilnadu, India.
 

Valliapan, R., Sumari, P., (Sept. 2010), “A Theoretical Methodology for Detection Segmentation Classification of Digital Mammogram Tumor by Machine Learning and Problem Solving Approach” International Journal of Computer Science Issues (IJCSI), Vol. 7, Issue 5