Abdulwahhab, Ali H.Bayat, OğuzIbrahim, Abdullahi Abdu2025-06-112025-06-112025Abdulwahhab, A. H., Bayat, O., & Ibrahim, A. A. (2025). HAFMAB-Net: hierarchical adaptive fusion based on multilevel attention-enhanced bottleneck neural network for breast histopathological cancer classification. Signal, Image and Video Processing, 19(5), 410.1863-17031863-1711https://hdl.handle.net/20.500.12939/5766Histological images play a crucial role in diagnosing diseases, especially breast cancer, which remains a major health concern for women worldwide. Computer-aided diagnosis tools significantly assist physicians in early detection and treatment planning, helping reduce mortality rates. Convolutional neural networks (CNNs) based on deep learning have proven effective in distinguishing benign from malignant breast cancers. In this context, HAFMAB-Net: Hierarchical Adaptive Fusion based on Multilevel Attention-Enhanced Bottleneck Neural Network, is proposed. The network comprises two pathways utilizing an enhanced Bottleneck architecture with attention mechanisms to extract both global and spatial features. It incorporates a Deeper Spatial Attention Aggregator Module to boost the representation of locative features by focusing on key spatial regions, improving the discriminative power of aggregated features. Additionally, a modified Adaptive Fusion Module combines the enhanced global and boosted spatial features into a comprehensive and enriched feature representation, which is subsequently used for classification. The proposed HAFMAB-Net was evaluated on the BACH dataset and further tested on the BreaKHis and LC25000 datasets to validate its robustness. The model achieved 99% accuracy on the BACH dataset, 98.99% accuracy on BreaKHis, 100% accuracy on each Colon, Lung, LC25000 datasets, respectively. These results highlight the HAFMAB-Net's efficiency, accuracy, and effectiveness in both multi-class and binary classification tasks, demonstrating its potential for broader applications in medical image analysis.eninfo:eu-repo/semantics/closedAccessAttention mechanismBottleneck neural networkBreast histopathological imageClassificationDeep learningArticle10.1007/s11760-025-04001-11952-s2.0-105000228482Q2WOS:001449039600001Q3