AUTHOR=A Ahila , M Poongodi , Bourouis Sami , Band Shahab S. , Mosavi Amir , Agrawal Shweta , Hamdi Mounir 

TITLE=Meta-Heuristic Algorithm-Tuned Neural Network for Breast Cancer Diagnosis Using Ultrasound Images

JOURNAL=Frontiers in Oncology

VOLUME=Volume 12 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2022.834028

DOI=10.3389/fonc.2022.834028

ISSN=2234-943X

ABSTRACT=Breast cancer is the most menacing cancers among all types of cancers in women around the globe. Early diagnosis is the only way to increase the treatment options which decreases the death rate and increases the chance of survival of patients. However, it is a big challenging task to differentiate abnormal breast tissues from normal ones because of their structure and unclear boundaries. Therefore, early and accurate diagnosis and classification of breast lesion into malignant or benign lesion is an active domain of research. Over the decade, numerous  Artificial Neural Network (ANN) based techniques were adopted in order to diagnose and classify the breast cancer due to their unique characteristics of learning key features from complex data via training process. However, these schemes have limitations like slow convergence and require more training time. To address the above mentioned issues, this paper employs metaheuristic algorithm for tuning the parameters of the neural network. A computer aided diagnosis scheme has been proposed for detecting abnormalities in the breast ultrasound images by integrating Wavelet Neural Network (WNN) and Grey Wolf Optimization (GWO) algorithm. Here, breast Ultrasound (US) images are preprocessed with sigmoid filter followed by interference based despeckling and then by An isotropic Diffusion. Automatic segmentation algorithm adopted to extract the region of interest and subsequently morphological and texture features were computed. Finally, GWO tuned WNN was exploited to accomplish the classification task. Classification performance of the proposed scheme is validated on 346 ultrasound images. Efficiency of the proposed methodology is evaluated by computing confusion matrix and Receiver Operating Characteristic (ROC) curve. Numerical analysis revealed that the proposed work can yield higher classification accuracy when compared to the prevailing methods and thereby proves its potential in effective breast tumor detection and classification.