Models & Optimisation and Mathematical Analysis Journal
Volume 8, Numéro 1, Pages 22-27
2020-12-31
Authors : Merati Medjeded . Mahmoudi Saïd . Chenine Abdelkader . Chikh Mohamed Amine .
In this paper, we propose a computer-aided diagnostic (CAD) system for breast images based on a statistical characterization combined to a classifier method. In the module of computer-aided detection (CADe), the co-ocurrences matrix was applied to the whole of image, while in the module of computer-aided identification (CADx), the same matrix was calculated only for the region of interest (ROI). For the both modules, the features are passed as inputs to the Learning Vector Quantization neural network (LVQ-NN) in order to classify the images in first module at normal or abnormal cases and at benign or malign cases in the second module. Our system was tested by using two samples of breast images composed respectively of two known bases of mammographic images namely: Digital Database for Screening Mammography (DDSM)[19] and mini-MIAS base generated by Mammographic Image Analysis Society[20]. A correct classification rate of 85.71% has been achieved by the CADe module and of 84.61% has been realized by the CADx module.
Breast images ; Computer-aided diagnostic - CAD ; Co-occurrences matrix ; Learning Vector Quantization - LVQ
Sabek Amine
.
Sََaihi Youcef
.
pages 475-492.
Hadjira Maouz
.
Asma Adda
.
Salah Hanini
.
pages 45-52.
عينوس رضوان
.
ص 333-352.
Gourine B
.
pages 139-160.