Transfer Learning and Handcrafted Features Ensembles for Ultrasound Breast Cancer Image Classification

Vanessa Kaplum Foleis; Bianca Aparecida Andrade; Henrique Bergamo Shigihara; Dimas Augusto Mendes Lemes; José Guilherme Picolo; Bernardo Feijó Junqueira; Guilherme Ribeiro Sales; Valentino Corso; Cides S. Bezerra

doi:10.22456/2175-2745.143357

Authors

Vanessa Kaplum Foleis CPQD - Centro de Pesquisa e Desenvolvimento
Bianca Aparecida Andrade CPQD - Centro de Pesquisa e Desenvolvimento https://orcid.org/0009-0006-2346-976X
Henrique Bergamo Shigihara CPQD - Centro de Pesquisa e Desenvolvimento https://orcid.org/0009-0001-1448-8225
Dimas Augusto Mendes Lemes Pontifícia Universidade Católica de Campinas (PUCC)
José Guilherme Picolo Pontifícia Universidade Católica de Campinas (PUCC) https://orcid.org/0009-0002-0214-4932
Bernardo Feijó Junqueira CPQD - Centro de Pesquisa e Desenvolvimento
Guilherme Ribeiro Sales CPQD - Centro de Pesquisa e Desenvolvimento
Valentino Corso CPQD - Centro de Pesquisa e Desenvolvimento https://orcid.org/0009-0004-5253-5151
Cides S. Bezerra CPQD - Centro de Pesquisa e Desenvolvimento

DOI:

https://doi.org/10.22456/2175-2745.143357

Keywords:

Breast cancer, ultrasound images, classification, transfer learning, late-fusion ensemble, cross-dataset analysis

Abstract

Breast cancer is the most commonly diagnosed cancer in women. Its diagnosis via ultrasound imaging largely depends on the technical skill of the radiologist. This study developed a binary classification system for breast lesions, combining transfer learning models and handcrafted features in ultrasound images. Pre-trained CNNs like InceptionV3, EfficientNetB4, ResNet50, and VGG16 were used, along with SVM-classified handcrafted features. Models were individually analyzed and combined using late-fusion ensembles. ResNet50 achieved an F1-score of 81.97%. The best late-fusion ensemble model reached an F1-score of 83.90%. In the cross-dataset evaluation, the top late-fusion ensemble model in the development dataset scored an F1-score of 88.70% and 78.20% in the test BUSI and BUID datasets, respectively. These results emphasize the robust potential of a late-fusion ensemble that combines CNN transfer learning and handcrafted features to classify breast lesions in ultrasound images.

Downloads

Download data is not yet available.

References

ŁUKASIEWICZ, S. et al. Breast cancer—epidemiology, risk factors, classification, prognostic markers, and current treatment strategies—an updated review. Cancers, v. 13, n. 17, p. 4287, 2021.

WANG, Y.; YAO, Y. Breast lesion detection using an anchor-free network from ultrasound images with segmentation-based enhancement. Scientific Reports, v. 12, n. 1, p. 14720, 2022.

HOOLEY, R. J.; SCOUTT, L. M.; PHILPOTTS, L. E. Breast ultrasonography: State of the art. Radiology, v. 268, n. 3, p. 642–659, 2013.

ISLAM, M. T.; TASCIOTTI, E.; RIGHETTI, R. Estimation of vascular permeability in irregularly shaped cancers using ultrasound poroelastography. IEEE Transactions on Biomedical Engineering, v. 67, n. 4, p. 1083–1096, 2020.

GREENSPAN, H.; GINNEKEN, B. van; SUMMERS, R. M. Guest editorial deep learning in medical imaging: Overview and future promise of an exciting new technique. IEEE Transactions on Medical Imaging, v. 35, n. 5, p. 1153–1159, 2016.

AKKUS, Z. et al. A survey of deep-learning applications in ultrasound: Artificial intelligence-powered ultrasound for improving clinical workflow. Journal of the American College of Radiology, v. 16, n. 19, p. 1546–1440, 2019.

ZOURHRI, M. et al. Deep learning technique for classification of breast cancer using ultrasound images. In: 2023 3rd International Conference on Innovative Research in Applied Science, Engineering and Technology (IRASET). Mohammedia, Morocco: IEEE, 2023. p. 1–8.

ARDABILI, S.; MOSAVI, A.; FELDE, I. Advances of deep learning in breast cancer modeling. In: 2023 IEEE 21st Jubilee International Symposium on Intelligent Systems and Informatics (SISY). Pula, Croatia: IEEE, 2023. p. 501–508.

CRUZ-RAMOS, C. et al. Benign and malignant breast tumor classification in ultrasound and mammography images via fusion of deep learning and handcraft features. Entropy, v. 25, n. 7, p. 991, 2023.

AL-DHABYANI, W. et al. Dataset of breast ultrasound images. Data in Brief, v. 28, p. 104863, 2020.

ARDAKANI, A. A. et al. An open-access breast lesion ultrasound image database: Applicable in artificial intelligence studies. Computers in Biology and Medicine, v. 152, p. 106438, 2023.

ALI, M. D. et al. Breast cancer classification through meta-learning ensemble technique using convolution neural networks. Diagnostics, v. 13, n. 13, p. 2242, 2023.

MOON, W. K. et al. Computer-aided diagnosis of breast ultrasound images using ensemble learning from convolutional neural networks. Computer Methods and Programs in Biomedicine, v. 190, p. 105361, 2020.

ABHISHEKA, B. et al. Combining handcrafted and CNN features for robust breast cancer detection using ultrasound images. In: 2023 3rd International Conference on Innovative Sustainable Computational Technologies (CISCT). New Delhi, India: IEEE, 2023. p. 1–6.

ABHISHEKA, B. et al. The efficacy of combining deep and handcrafted features for breast cancer classification using ultrasound images. In: 2023 10th IEEE Uttar Pradesh Section International Conference on Electrical, Electronics and Computer Engineering (UPCON). India: IEEE, 2023. v. 10, p. 735–740.

NANNI, L. et al. Digital recognition of breast cancer using TakhisisNet. In: [S.l.]: IGI Global, 2022. p. 1286–1304.

MO, Y. et al. Hover-Trans: Anatomy-aware Hover-Transformer for ROI-free breast cancer diagnosis in ultrasound images. IEEE Transactions on Medical Imaging, v. 42, n. 6, p. 1696–1706, 2023.

GÓMEZ-FLORES, W.; GREGORIO-CALAS, M. J.; PEREIRA, W. C. D. A. BUS-BRA: A breast ultrasound dataset for assessing computer-aided diagnosis systems. Medical Physics, p. 1–14, 2023.

MAYERHOEFER, M. E. et al. Introduction to radiomics. Journal of Nuclear Medicine, v. 61, n. 4, p. 488–495, 2020.

DHEEPAK, G.; J., A. C.; VAISHALI, D. Brain tumor classification: A novel approach integrating GLCM, LBP and composite features. Frontiers in Oncology, v. 13, p. 1248452, 2024.

HALL-BEYER, M. GLCM texture: A tutorial. National Council on Geographic Information and Analysis Remote Sensing Core Curriculum, University of Calgary Calgary, AB, Canada, v. 3, n. 1, p. 75, 2000.

OJALA, T.; PIETIKAINEN, M.; MAENPAA, T. Multiresolution gray-scale and rotation invariant texture classification with local binary patterns. IEEE Transactions on Pattern Analysis and Machine Intelligence, v. 24, n. 7, p. 971–987, 2002.

OJANSIVU, V.; HEIKKILÄ, J. Blur insensitive texture classification using local phase quantization. Image and Signal Processing, Springer Berlin Heidelberg, Berlin, Heidelberg, p. 236–243, 2008.

RAO, K. S. et al. Intelligent ultrasound imaging for enhanced breast cancer diagnosis: Ensemble transfer learning strategies. IEEE Access, v. 12, p. 22243–22263, 2024.

KITTLER, J. et al. On combining classifiers. IEEE Transactions on Pattern Analysis and Machine Intelligence, v. 20, n. 3, p. 226–239, 1998.

GESNOUIN, J. et al. Assessing cross-dataset generalization of pedestrian crossing predictors. In: 2022 IEEE Intelligent Vehicles Symposium (IV). [S.l.: s.n.], 2022. p. 419–426.