A novel bat optimized runlength networks (BORN) for an efficient classification of breast cancer

In recent times, Medical Information and Processing deals with various methodologies utilized for the prognosis and diagnosis of various harmful diseases with the help of trending artificial intelligence and machine learning techniques. Breast cancer is one of such disease which occupies the major share in killing the millions of people especially women. Several intelligent methods were proposed for an efficient diagnosis of breast cancer, but brighter light of research is required for better diagnosis. Hence the new methodology of integrating the run length features along with the Bat optimized learning Machines-BORN has been proposed. BORN also features the most efficient visual saliency segmentation process to obtain highly efficient diagnosis. The main aim of the proposed BORN algorithm is to diagnosis the different stages of breast cancer with high accuracy and minimal error. For attaining the high accuracy, BORN has been tested with two different datasets MIAS and DDSM with different learning kernels and compared with the other intelligent algorithms such as RF-ELM, EGAM and Associate Classifiers in which accuracy of the proposed classifier is 99.5%.

Automated summary

This study proposes a new methodology that integrates run length features with Bat Optimized Learning Machines for improved diagnosis of breast cancer. Through testing on different datasets and comparison with other intelligent algorithms, the proposed classifier achieved an accuracy of 99.5%.