The Ensemble technique uses variable hidden neurons hierarchical fusion and tenfold cross validation for Detection of Breast masses

Chandra Bose, Subash and Babu Changalasetty, Suresh and Said Badawy, Ahmed and Ghribi, Wade and Marzougui, Mehrez and Saroja Thota, Lalitha (2016) The Ensemble technique uses variable hidden neurons hierarchical fusion and tenfold cross validation for Detection of Breast masses. Biohouse Journal of Computer Science, 2 (2). pp. 12-18. ISSN ISSN 2379-1500

[img] PDF
BJCS-V2-I2-4.pdf

Download (375kB)
Official URL: http://biohouse.us/docs/BJCS-V2-I2-4.pdf

Abstract

Early stage detection of breast cancer using image processing methods in digital mammograms is the best method. Enhancement of the variable hidden neural network for detecting the spot of breast mass is presented. It can be done in several stages, first the preprocessing methods have been done over digital mammogram image. Second stage the Region of Interest - ROI is extracted from the processed image. Third stage the region Growing Segmentation is implemented to separate the part of the image that having the identical pixel values from the mammogram. Fourth stage the features extracted such as density, mass shape, mass margin, patient age, subtlety, and abnormality assessment rank value. Fifth the neural networks by varying the number of neurons in the hidden layer. According to the classification of accuracy these are then trained, tested and ranked. To create an ensemble network, the Ten-Fold cross validation which produces the classifiers, is used. The Artificial neural network classifiers are then fused together using majority vote algorithm to create the final ensemble network which reveals whether the image is malignant or benign. The classification is done using the Digital Database for Screening Mammography (DDSM).The DDSM is organized into “cases” and “volumes”. A “case” is a collection of images and information corresponding to one mammography exam of one patient. A “volume” is simply a collection of cases collected together for the purpose of ease of distribution. This database contains 2620 cases in 43 volumes. The Experimental results were discussed using MATLAB.

Item Type: Article
Subjects: Computer Engineering
Computer Sciences
Divisions: College of Computer Sciences > Computer Engineering
Depositing User: subash jaganathan
Date Deposited: 11 Mar 2017 13:04
Last Modified: 11 Mar 2017 13:04
URI: http://eprints.kku.edu.sa/id/eprint/534

Actions (login required)

View Item View Item