@article{MAKHILLJEAS202015419023,
    title = {Employing Symmetry Concept and Unsupervised Neural Network to
Detect Abnormal Regions in IR Breast Thermograms},
    journal = {Journal of Engineering and Applied Sciences},
    volume = {15},
    number = {4},
    pages = {898-907},
    year = {2020},
    issn = {1816-949x},
    doi = {jeasci.2020.898.907},
    url = {https://makhillpublications.co/view-article.php?issn=1816-949x&doi=jeasci.2020.898.907},
    author = {Rabab Saadoon},
    keywords = {unsupervised neural network,Symmetry,IR thermograms,breast cancer,infrared,thermography},
    abstract = {Infrared thermography is one of many medical scanning for detecting breast cancer and other
abnormalities. Women breasts possess a high degree of symmetry, this property is employed in this research
to detect the presence of abnormalities in breast front view thermograms of left and right breasts by proposing
symmetry line algorithm. Clustering process utilizing unsupervised Self Organization Feature Map (SOFM)
neural network is a second proposed technique in this work to isolate and extract abnormal regions in IR breast
thermograms. The results declared the efficiency of the proposed symmetry line algorithm, depending on the
histogram and standard deviation calculated values to detect breast abnormalities in the experimental
thermograms. As well as the results of the second proposed unsupervised neural network clustering method
proved its effective and adequate performance, it succeeded to extract the cancer and other abnormal regions
in the abnormal sets of breast thermograms.}
    }