Fully automated scheme for computer-aided detection and breast cancer diagnosis using digitised mammograms

• Published in: IET image processing Vol. 14; no. 3; pp. 495 - 505
• Main Authors: Eltrass, Ahmed S, Salama, Mohamed S
• Format: Journal Article
• Language: English
• Published: The Institution of Engineering and Technology 28.02.2020
• Subjects: benign–malignant classification; benign–malignant patterns; breast cance early detection; breast cancer diagnosis; CAD system; cancer; computer‐aided detection; correlation bias reduction algorithm; correlation methods; digital mammograms; expectation‐maximisation algorithm; feature extraction; feature extraction technique; feature selection; fully automated computer‐aided diagnosis system; hybrid feature selection approach; image classification; image resolution; image texture; learning (artificial intelligence); low‐computational requirements; mammography; medical image processing; multiorientation features; multiresolution features; object detection; public health problem; Q similarity‐based learning algorithm; real clinical mammograms; real‐time systems; region of interest extraction; Research Article; ROIs; statistical analysis; support vector machine recursive feature elimination; support vector machines; textural features; time 2.2 s; wavelet transforms; wavelet‐based contourlet transform; digital mammograms; benign–malignant classification; wavelet transforms; image classification; multiorientation features; breast cancer diagnosis; fully automated computer-aided diagnosis system; time 2.2 s; object detection; computer-aided detection; region of interest extraction; multiresolution features; feature extraction technique; feature extraction; wavelet-based contourlet transform; Q similarity-based learning algorithm; learning (artificial intelligence); real clinical mammograms; medical image processing; image resolution; feature selection; ROIs; support vector machines; mammography; CAD system; hybrid feature selection approach; correlation bias reduction algorithm; image texture; support vector machine recursive feature elimination; expectation–maximisation algorithm; low-computational requirements; benign–malignant patterns; textural features; public health problem; real-time systems; expectation-maximisation algorithm; cancer; statistical analysis; breast cance early detection; correlation methods
• ISSN: 1751-9659; 1751-9667; 1751-9667
• DOI: 10.1049/iet-ipr.2018.5953

Breast cancer becomes a significant public health problem in the world. During the early detection of breast cancer, it is a very challenging task to classify accurately the benign–malignant patterns in digital mammograms. This study proposes a new fully automated computer-aided diagnosis (CAD) system for breast cancer diagnosis with high-accuracy and low-computational requirements. The expectation–maximisation algorithm is investigated to extract automatically the region of interests (ROIs) within mammograms. The standard shape, statistical, and textural features of ROIs are extracted and combined with multi-resolution and multi-orientation features derived from a new feature extraction technique based on wavelet-based contourlet transform. A hybrid feature selection approach based on combining the support vector machine recursive feature elimination with correlation bias reduction algorithm is proposed. Also, the authors investigate a new similarity-based learning algorithm, called Q, for benign–malignant classification. The proposed CAD system is applied to real clinical mammograms, and the experimental results demonstrate the superior performance of the proposed CAD system over other existing CAD systems in terms of accuracy 98.16%, sensitivity 98.63%, specificity 97.80%, and computational time 2.2 s. This reveals the effectiveness of the proposed CAD system in improving the accuracy of breast cancer diagnosis in real-time systems.