FussCyier: Mamogram images classification based on similarity measure fuzzy soft set

• Main Authors: Lashari, Saima Anwar, Ibrahim, Rosziati, Senan, Norhalina
• Format: Conference or Workshop Item
• Language: English
• Published: 2017
• Subjects: QA75 Electronic computers. Computer science
• Online Access: http://repo.uum.edu.my/22795/1/ICOCI%202017%2056-61.pdf; http://repo.uum.edu.my/22795/; http://icoci.cms.net.my/PROCEEDINGS/2017/Pdf_Version_Chap01e/PID186-56-61e.pdf

Automatic digital mammograms reading become highly enviable, as the number of mammograms to be examined by physician increases enormously.It is premised that the computer aided diagnosis system is mandatory to assist physicians/radiologists to achieve high efficiency and productivity.To handle uncertainties of medical images, fuzzy soft set theory has been merely scrutinized, even though the choice of convenient parameterization makes fuzzy soft set suitable and feasible for decision making applications. Therefore, this study investigates the practicability of fuzzy soft set for classification of digital mammogram images to increase the classification accuracy while lower the classifier complexity.The proposed method FussCyier involves three phases namely: pre-processing, training and testing.Results of the research indicated that proposed method gives high classification performance with wavelet de-noise filter Sym8 with the accuracy 75.64%, recall 84.67% and CPU time 0.0026 seconds.