Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{205339,
        author = {Mrs. Y. Sheela and Ms. M. Sivasankari},
        title = {ATTENTION-BASED EXPLAINABLE DEEP LEARNING FRAMEWORK FOR FETAL ULTRASOUND IMAGE PLANE CLASSIFICATION},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {13},
        number = {1},
        pages = {6173-6180},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=205339},
        abstract = {Fetal ultrasound imaging is an essential non-invasive diagnostic technique used for monitoring fetal growth, anatomical development, and prenatal health assessment. Accurate identification of standard ultrasound planes is crucial for detecting abnormalities and supporting clinical decision-making. However, manual classification of fetal ultrasound images is challenging due to speckle noise, low image contrast, artifacts, fetal movement, and operator dependency. These limitations may lead to inconsistent diagnoses and increased workload for healthcare professionals. This project proposes an Attention-Based Explainable Deep Learning Framework for Fetal Ultrasound Image Plane Classification. The proposed framework utilizes DenseNet121 as the backbone network for automatic feature extraction and image classification. To enhance the model’s ability to focus on clinically significant anatomical structures, a Convolutional Block Attention Module (CBAM) is integrated into the network architecture. The attention mechanism improves feature representation by emphasizing important spatial and channel-wise information while suppressing irrelevant background features. The system classifies fetal ultrasound images into six categories: Fetal Brain, Fetal Abdomen, Fetal Thorax, Fetal Femur, Maternal Cervix, and Others. Furthermore, Explainable Artificial Intelligence (XAI) is incorporated through Gradient-weighted Class Activation Mapping (Grad-CAM), which generates visual explanations highlighting the regions responsible for classification decisions. This improves model transparency and increases clinicians’ confidence in the automated predictions. Experimental evaluation demonstrates that the proposed attention-based framework achieves improved classification performance in terms of accuracy, precision, recall, and F1-score compared with conventional convolutional neural network models. The integration of attention mechanisms and explainability techniques provides both high predictive performance and interpretability. Therefore, the proposed framework can serve as an effective computer-aided diagnostic tool to assist radiologists and sonographers in fetal ultrasound examinations, ultimately improving prenatal healthcare and diagnostic reliability.},
        keywords = {Fetal Ultrasound Imaging, Plane Classification, DenseNet121, CBAM, Deep Learning, Explainable Artificial Intelligence, Grad-CAM, Medical Image Analysis.},
        month = {June},
        }