RETRACTED ARTICLE: Region-specific multi-attribute white mass estimation-based mammogram classification

The problem of mammographic image classification has been handled using various measures and features. The methods consider only small set of features to perform classification, but still the methods suffer to produce efficient classification accuracy. To overcome the problem of accuracy in mammogra...

Full description

Saved in:
Bibliographic Details
Published inPersonal and ubiquitous computing Vol. 22; no. 5-6; pp. 1093 - 1098
Main Authors Padmavathy, T. V., Vimalkumar, M. N., Sivakumar, N.
Format Journal Article
LanguageEnglish
Published London Springer London 01.10.2018
Springer Nature B.V
Subjects
Computer Science
Mobile Computing
Original Article
Personal Computing
User Interfaces and Human Computer Interaction
Mammogram classification
Peak white
Region-based classification
White density
White mass value
Online AccessGet full text

Cover

More Information
Summary:The problem of mammographic image classification has been handled using various measures and features. The methods consider only small set of features to perform classification, but still the methods suffer to produce efficient classification accuracy. To overcome the problem of accuracy in mammographic image classification, a region-specific multi-attribute white mass estimation technique is proposed. The method uses the white mass value, density measure, and binding to identify the microcalcification. First, the peak white mass value is identified by visiting throughout the mammogram region. Second, the method splits the mammographic image into a number of small scale integral images. Third, for each integral image, the method computes multi-attribute white mass value, and based on computed white mass value, the method identifies the region being affected by the calcification. The method produces efficient result in mammogram image classification.
ISSN:1617-4909
1617-4917
DOI:10.1007/s00779-018-1135-4