An optimized ensemble model based on meta-heuristic algorithms for effective detection and classification of breast tumors

• DOI: 10.1007/s00521-024-10719-9
• Publication Date: 2024-12-27T05:50:15Z
• Authors: Abeer Saber; Samar Elbedwehy; Wael A. Awad; Esraa Hassan
• Keywords: 03 medical and health sciences; 0302 clinical medicine

Abstract One of the most common cancers among women worldwide is breast cancer (BC), and early diagnosis can save lives. Early detection BC increases likelihood a successful outcome by enabling treatment to start sooner. Even in areas without access specialist physician, machine learning (ML) aids detection. The medical imaging community becoming more interested using ML, deep (DL) increase accuracy screening. Many disease-related data are sparse. However, for DL models perform well, large amount required. Because this, that currently use on images not as effective they could be. Convolutional neural network (CNN) have recently gained popularity industry, admirably terms high performance robustness at image classification. proposed method classifies ensemble pre-trained such dense convolutional (DenseNet)-121 EfficientNet-B5 feature extractor networks, well support vector Using modified meta-heuristic optimizer, selected CNN hyperparameters were optimized improve performance. experimental results presented model INbreast dataset show classification, with overall accuracy, sensitivity, specificity, precision, area under ROC curve (AUC) values 99.9%, 99.8%, 99.1%, 1.0, respectively.