Citation :

Nur Athiqah Harron, Siti Noraini Sulaiman, Muhammad Khusairi Osman, Iza Sazanita Isa, Noor Khairiah A. Karim, Slamet Riyadi, "CNNSVM-BF: A Robust Hybrid Feature-Based Approach for Blur Detection in Digital Breast Tomosynthesis Images," International Journal of Electrical and Electronics Engineering, vol. 13, no. 5, pp. 11-22, 2026. Crossref, https://doi.org/10.14445/23488379/IJEEE-V13I5P102

Abstract

Blur detection is an essential yet challenging task in digital image processing, particularly in medical imaging applications where image quality directly affects analysis reliability. In Digital Breast Tomosynthesis (DBT), motion artifacts and limited-angle acquisition frequently cause image blurring, degrading fine structural details, and complicating automated analysis. Blur detection in DBT is further challenged by the lack of prior blur knowledge and the visual similarity between blurred and sharp regions. This paper proposes a hybrid feature-based CNNSVM approach for detecting blur in DBT images by jointly utilizing automated image features extracted from a CNN with a BF (Laplacian-based Blur Factor). The BF, based on the variance of the Laplacian response, is applied to measure edge degradation and is used in conjunction with CNN-based features to enhance classification performance. The integrated feature set is characterized by a Support Vector Machine (SVM) model to identify blurred or non-blurred DBT images. Experimental evaluation using a large publicly available DBT dataset indicates that the hybrid scheme achieves an accuracy of 99.21% in blur detection. The model performed much better than CNN and CNNSVM models that rely only on image features. Furthermore, the hybrid design reduces complexity and maintains good performance at reduced training times. These findings indicate that the proposed framework provides a practical way to automatically blur detection in medical image processing tasks.

Keywords

Blur detection, Convolutional Neural Network, Digital Breast Tomosynthesis, Hybrid features, Laplacian operator.

References

[1]   Ji Soo Choi et al., “Comparison of Synthetic and Digital Mammography with Digital Breast Tomosynthesis or Alone for the Detection and Classification of Microcalcifications,” European Radiology, vol. 29, no. 1, pp. 319-329, 2019.
[CrossRef] [Google Scholar] [Publisher Link]

[2]   Parvinder S. Sujlana et al., “Digital Breast Tomosynthesis: Image Acquisition Principles and Artifacts,” Clinical Imaging, vol. 55, pp. 188-195, 2019.
[CrossRef] [Google Scholar] [Publisher Link]

[3]   Pranjal Sahu et al., “Using Virtual Digital Breast Tomosynthesis for De-Noising of Low-Dose Projection Images,” 2019 IEEE 16^th International Symposium on Biomedical Imaging (ISBI 2019), Venice, Italy, pp. 1647-1651, 2019.
[CrossRef] [Google Scholar] [Publisher Link]

[4]   Joana Boita et al., “How Does Image Quality Affect Radiologists’ Perceived Ability for Image Interpretation and Lesion Detection in Digital Mammography?,” European Radiology, vol. 31, no. 7, pp. 5335-5343, 2021.
[CrossRef] [Google Scholar] [Publisher Link]

[5]   Amar Kavuri, and Mini Das, “Relative Contributions of Anatomical and Quantum Noise in Signal Detection and Perception of Tomographic Digital Breast Images,” IEEE Transactions on Medical Imaging, vol. 39, no. 11, pp. 3321-3330, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[6]   Ahmed K. Abdullah et al., “The Impact of Simulated Motion Blur on Lesion Detection Performance in Full-Field Digital Mammography,” British Journal of Radiology, vol. 90, no. 1075, pp. 1-7, 2017.
[CrossRef] [Google Scholar] [Publisher Link]

[7]   Christoph I. Lee, and Constance D. Lehman, “Digital Breast Tomosynthesis and the Challenges of Implementing an Emerging Breast Cancer Screening Technology Into Clinical Practice,” Journal of the American College of Radiology, vol. 13, no. 11, pp. R61-R66, 2016.
[CrossRef] [Google Scholar] [Publisher Link]

[8]   Pierpaolo Pattacini et al., “Digital Mammography Versus Digital Mammography Plus Tomosynthesis for Breast Cancer Screening: The Reggio Emilia Tomosynthesis Randomized Trial,” Radiology, vol. 288, no. 2, pp. 375-385, 2018.
[CrossRef] [Google Scholar] [Publisher Link]

[9]   Akshat C. Pujara et al., “Digital Breast Tomosynthesis Slab Thickness: Impact on Reader Performance and Interpretation Time,” Radiology, vol. 297, no. 3, pp. 534-542, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[10] Awais Khan et al., “A Robust Approach for Blur and Sharp Regions’ Detection Using Multisequential Deviated Patterns,” International Journal of Optics, vol. 2021, pp. 1-13, 2021.
[CrossRef] [Google Scholar] [Publisher Link]

[11] Muhammad Tariq Mahmood, Usman Ali, and Young Kyu Choi, “Single Image Defocus Blur Segmentation Using Local Ternary Pattern,” ICT Express, vol. 6, no. 2, pp. 113-116, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[12] Awais Khan et al., “Defocus Blur Detection using Novel Local Directional Mean Patterns (LDMP) and Segmentation via KNN Matting,” Frontiers of Computer Science, vol. 16, no. 2, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[13] Magudeeswaran Veluchamy, and Bharath Subramani, “Fuzzy Dissimilarity Color Histogram Equalization for Contrast Enhancement and Color Correction,” Applied Soft Computing, vol. 89, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[14] Taohong Zhang et al., “A Feature Fusion Method with Guided Training for Classification Tasks,” Computational Intelligence and Neuroscience, vol. 2021, no. 1, 2021.
[CrossRef] [Google Scholar] [Publisher Link]

[15] A. Maldera et al., “Digital Breast Tomosynthesis: Dose and Image Quality Assessment,” Physica Medica, vol. 33, pp. 56-67, 2017.
[CrossRef] [Google Scholar] [Publisher Link]

[16] Clement Farabet et al., “Learning Hierarchical Features for Scene Labeling,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 35, no. 8, pp. 1915-1929, 2013.
[CrossRef] [Google Scholar] [Publisher Link]

[17] Roxanne A. Pagaduan, Ma. Christina R. Aragon, and Ruji P. Medina, “iBlurDetect: Image Blur Detection Techniques Assessment and Evaluation Study,” Proceedings of the International Conference on Culture Heritage, Education, Sustainable Tourism, and Innovation Technologies (CESIT 2020), vol. 1, pp. 286-291, 2020.
[Google Scholar] [Publisher Link]

[18] Ali Karaali, and Claudio Rosito Jung, “Edge-based Defocus Blur Estimation with Adaptive Scale Selection,” IEEE Transactions on Image Processing, vol. 27, no. 3, pp. 1126-1137, 2018.
[CrossRef] [Google Scholar] [Publisher Link]

[19] Guodong Xu, Yuhui Quan, and Hui Ji, “Estimating Defocus Blur via Rank of Local Patches,” Proceedings of the IEEE International Conference on Computer Vision (ICCV), Venice, Italy, pp. 5381-5389, 2017.
[CrossRef] [Google Scholar] [Publisher Link]

[20] S. Alireza Golestaneh, and Lina J. Karam, “Spatially-Varying Blur Detection based on Multiscale Fused and Sorted Transform Coefficients of Gradient Magnitudes,” 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 596-605, 2017.
[CrossRef] [Google Scholar] [Publisher Link]

[21] Chang Tang et al., “A Spectral and Spatial Approach of Coarse-to-Fine Blurred Image Region Detection,” IEEE Signal Processing Letters, vol. 23, no. 11, pp. 1652-1656, 2016.
[CrossRef] [Google Scholar] [Publisher Link]

[22] Bolan Su, Shijian Lu, and Chew Lim Tan, “Blurred Image Region Detection and Classification,” MM '11: Proceedings of the 19^th ACM International Conference on Multimedia, Scottsdale, Arizona, USA, pp. 1397-1400, 2011.
[CrossRef] [Google Scholar] [Publisher Link]

[23] Jinsun Park et al., “A Unified Approach of Multi-Scale Deep and Hand-Crafted Features for Defocus Estimation,” 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA, pp. 2760-2769, 2017.
[CrossRef] [Google Scholar] [Publisher Link]

[24] Rui Huang et al., “Multiscale Blur Detection by Learning Discriminative Deep Features,” Neurocomputing, vol. 285, pp. 154-166, 2018.
[CrossRef] [Google Scholar] [Publisher Link]

[25] Kai Zeng et al., “A Local Metric for Defocus Blur Detection Based on CNN Feature Learning,” IEEE Transactions on Image Processing, vol. 28, no. 5, pp. 2107-2115, 2018.
[CrossRef] [Google Scholar] [Publisher Link]

[26] Beomseok Kim et al., “Defocus and Motion Blur Detection with Deep Contextual Features,” Computer Graphics Forum, vol. 37, no. 7, pp. 277-288, 2018.
[CrossRef] [Google Scholar] [Publisher Link]

[27] Aodong Shen et al., “Automatic Extraction of Blur Regions on a Single Image Based on Semantic Segmentation,” IEEE Access, vol. 8, pp. 44867-44878, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[28] Xuewei Wang et al., “Efficient Image Blur Detection via Hierarchical Edge Guidance and Region Complementation,” Complex & Intelligent Systems, vol. 9, no. 6, pp. 6523-6540, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[29] Tomasz Szandała, “Convolutional Neural Network for Blur Images Detection as an Alternative for Laplacian Method,” 2020 IEEE Symposium Series on Computational Intelligence (SSCI), Canberra, ACT, Australia, pp. 2901-2904, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[30] Carlos Silva, and Daniel Welfer, “A Novel Hybrid SVM-CNN Method for Extracting Characteristics and Classifying Cattle Branding,” Latin-American Journal of Computing, vol. 6, no. 1, pp. 9-16, 2019.
[Google Scholar] [Publisher Link]

[31] M. O. Khairandish et al., “A Hybrid CNN-SVM Threshold Segmentation Approach for Tumor Detection and Classification of MRI Brain Images,” IRBM, vol. 43, no. 4, pp. 290-299, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[32] Jing Liu et al., “A Convolutional Neural Network for Image Super-Resolution Using Internal Dataset,” IEEE Access, vol. 8, pp. 201055-201070, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[33] Mateusz Buda et al., “Detection of Masses and Architectural Distortions in Digital Breast Tomosynthesis: A Publicly Available Dataset of 5,060 Patients and a Deep Learning Baseline,” arXiv Preprint, pp. 1-14, 2020.
[CrossRef] [Google Scholar] [Publisher Link]