Author(s)

Alio Boubacar Goga^*, Harouna Naroua, Chaibou Kadri

Laboratoire d’Informatique Fondamentale et Appliquée-Sciences de l’Ingénierie (LIFA-SI), Département de Mathématiques et Informatique, Faculté des Sciences et Techniques, Université Abdou Moumouni, Niamey, Niger.

In many fields, particularly that of health, the diagnosis of diseases is a very difficult task to carry out. Therefore, early detection of diseases using artificial intelligence tools can be of paramount importance in the medical field. In this study, we proposed an intelligent system capable of performing diagnoses for radiologists. The support system is designed to evaluate mammographic images, thereby classifying normal and abnormal patients. The proposed method (DiagBC for Breast Cancer Diagnosis) combines two (2) intelligent unsupervised learning algorithms (the C-Means clustering algorithm and the Gaussian Mixture Model) for the segmentation of medical images and an algorithm for supervised learning (a modified DenseNet) for the diagnosis of breast images. Ultimately, a prototype of the proposed system was implemented for the Magori Polyclinic in Niamey (Niger) making it possible to diagnose (or classify) breast cancer into two (2) classes: the normal class and the abnormal class.

Image Diagnosis, Segmentation, DenseNet, Unsupervised Learning, Supervised Learning, Breast Cancer

Goga, A. , Naroua, H. and Kadri, C. (2024) A Hybrid Learning Algorithm for Breast Cancer Diagnosis. Journal of Intelligent Learning Systems and Applications, 16, 262-273. doi: 10.4236/jilsa.2024.163014.