Breast Cancer Detection Using Variational Mode Decomposition (VMD) and Weighted Bidirectional Extreme Learning Machine
Sukumar Joshi, Indu Sekhar Samanta, Subhasis Panda, Pravat Kumar Rout, Kunjabihari Swain · 2023
Breast cancer is a principal cause of mortality among women globally, and early identification and diagnosis can significantly increase patients' survival rates. This paper introduces a novel approach for detecting breast cancer using Variational Mode Decomposition (VMD) and a Weighted Bidirectional Extreme Learning Machine (WBELM). The proposed technique utilizes VMD to break mammographic images into multiple modes, each signifying distinct frequency constituents. Following this decomposition, the WBELM is deployed to classify these modes into benign or malignant categories. The method's efficacy is assessed using the Digital Database for Screening Mammography (DDSM), and the simulations demonstrate a commendable accuracy rate of 97.61 %, a sensitivity rate of 97.54 %, and a specificity rate of 97.68%.