Recent Advances in Extreme Learning-Based Approaches for Breast Cancer Diagnosis
Sarthak Padhi, Suvendu Rup · 2019
This paper reviews some of the notable recent findings in breast cancer detection for Computer-Aided Diagnosis (CADx) systems using Extreme Learning Machine (ELM) approaches. Digital mammogram classification using machine learning is a popular technique used for minimizing the risk of beast cancer. In this paper, a brief survey on the usage and appreciability of ELMs for breast cancer detection in CADx systems has been studied. The databases mostly preferred for training and testing of the ELM-based systems include the Mini-MIAS and DDSM databases among a select group of synthetic mammographic databases; they may also include actual medical images procured from supportive hospitals. In most of the reported results, it has been observed that the ELM-based approaches for breast cancer detection show superior results in terms of accurate classification as well as accurate segmentation. The articles studied in this review show how ELMs have been utilized in both a basic form as well as various modified versions.