Classification of masses in mammographic image using wavelet domain features and polynomial classifier

作者：

Highlights：

• We propose a system for texture analysis and classification of lesions in mammograms.

• We validate the method by quantitative evaluation.

• We also made a comparison with other artificial intelligence algorithms.

• The method allows classification with average value 0.98 of Az.

• The method is effective in 360 digitized mammograms from DDSM database.

摘要

•We propose a system for texture analysis and classification of lesions in mammograms.•We validate the method by quantitative evaluation.•We also made a comparison with other artificial intelligence algorithms.•The method allows classification with average value 0.98 of Az.•The method is effective in 360 digitized mammograms from DDSM database.

论文关键词：Texture analysis,Mammography,Polynomial classifier,Wavelet,CAD

论文评审过程：Available online 9 May 2013.

论文官网地址：https://doi.org/10.1016/j.eswa.2013.04.036