Front Inner Page - Volume 3 No.4 August 2016

  • » Back to Index

  • Title

    :

    A Study on Micro-calcifications In Mammograms Using M-Band Wavelet Transform and Support Vector Machine

    Authors

    :

    C.Suba1, Dr.K.Nirmala2

    Keywords

    :

    Digital Mammography, Microcalcification, Benign, Malignant, Wavelet Transform, Support Vector Machine, Feature Extraction, Classification, Energy Computation, MIAS Database.

    Issue Date

    :

    August 2016

    Abstract

    :

    Digital Mammography is a most important vital task early detection of breast cancer especially in the case of microcalcification. Microcalcification is tiny calcium less than 1/50 of an inch of a millimeter appears in mammogram. Wavelet transform is an important multiresolution analysis tool has been commonly applied to single or image analysis. In this paper, an efficient computerized system for the classification of Microcalcification Clusters (MC) in digital mammogram based on Dual Tree M-Band Wavelet Transform (DTMBWT) and Support Vector Machine (SVM) classifier. The DTMBWT has some unique properties that provide local, multi-scale and directional analysis.Intially, the given mammogram is decomposed by using DTMBWT at a predefined level of decomposition. From each sub band, the wavelet energy is calculated. These energies are used as features to classify given mammogram into benign or malignant. Experimental results show that the classification rate of the proposed system is evaluated by using classification accuracy in percentage based on DTMBWT.The proposed methodology system results achieves 91.83% good classification accuracy on Image Analysis Society(MIAS) database Images.

    Page(s)

    :

    1-6

    ISSN

    :

    2347- 4734

    Source

    :

    Vol. 3, No.4, August 2016

    Download

    :


  • » Back index