(Institut Teknologi Sepuluh Nopember, Indonesia)(2) * Nanik Suciati
(Institut Teknologi Sepuluh Nopember, Indonesia)*corresponding author
AbstractAccurate breast ultrasound (BUS) lesion segmentation is critical for early diagnosis but is challenged by image artifacts and the reliance of foundation models on manual prompting. Existing automated frameworks often lack robust fail-safe mechanisms, leading to missed diagnoses. To address this reliability gap, this study proposes a novel, fully automated hybrid segmentation framework that synergistically integrates three key components: (1) a recall-optimized YOLOv9 detector tailored to minimize clinical false negatives; (2) a MedSAM2 foundation model efficiently fine-tuned via Low-Rank Adaptation (LoRA) for ultrasound specifics; and (3) a statistical fallback mechanism that acts as a crucial safety net to recover spatial prompts during detection failures. Evaluated on the public BUSI dataset, the recall-dominant detection module achieved a Recall of 0.8238. Supported by this robust prompting and fallback strategy, the segmentation module achieved a Dice coefficient of 0.8818 and an IoU of 0.8113. By effectively integrating specialized detection with adaptive segmentation and a statistical fail-safe, the proposed pipeline offers a highly reliable automated approach for computer-aided screening systems.
KeywordsBreast Ultrasound Segmentation; Segment Anything Model 2; Automated Prompting; Recall Optimization; Low-Rank Adaptation;
|
DOIhttps://doi.org/10.26555/ijain.v12i2.2408 |
Article metricsAbstract views : 213 | PDF views : 3 |
Cite |
Full Text Download
|
References
[1] A. Simon and K. Robb, “Cancer: breast,” in Cambridge Handbook of Psychology, Health and Medicine, Cambridge University Press, 2001, pp. 577–580. doi: 10.1017/CBO9780511543579.131.
[2] J. S. Ahn et al., “Artificial Intelligence in Breast Cancer Diagnosis and Personalized Medicine,” J. Breast Cancer, vol. 26, no. 5, p. 405, Oct. 2023, doi: 10.4048/jbc.2023.26.e45.
[3] Y. Shin et al., “Breast Cancer Risk After Hysterectomy: A Health Insurance Database-Based Analysis,” J. Breast Cancer, vol. 28, no. 4, p. 215, Aug. 2025, doi: 10.4048/jbc.2025.0031.
[4] M. A. Aslam, A. Naveed, N. Ahmed, and Z. Ke, “A hybrid attention network for accurate breast tumor segmentation in ultrasound images,” Sci. Rep., vol. 15, no. 1, p. 39633, Nov. 2025, doi: 10.1038/s41598-025-23213-6.
[5] G. Zhao, X. Zhu, X. Wang, F. Yan, and M. Guo, “Syn-Net: A Synchronous Frequency-Perception Fusion Network for Breast Tumor Segmentation in Ultrasound Images,” IEEE J. Biomed. Heal. Informatics, vol. 29, no. 3, pp. 2113–2124, Mar. 2025, doi: 10.1109/JBHI.2024.3514134.
[6] J. Huang et al., “EMGANet: Edge-Aware Multi-Scale Group-Mix Attention Network for Breast Cancer Ultrasound Image Segmentation,” IEEE J. Biomed. Heal. Informatics, vol. 29, no. 8, pp. 5631–5641, Aug. 2025, doi: 10.1109/JBHI.2025.3546345.
[7] F. A. Hermawati, H. Tjandrasa, and N. Suciati, “Phase-based thresholding schemes for segmentation of fetal thigh cross-sectional region in ultrasound images,” J. King Saud Univ. - Comput. Inf. Sci., vol. 34, no. 7, pp. 4448–4460, Jul. 2022, doi: 10.1016/j.jksuci.2021.02.004.
[8] M. H. A. M. H. Himel, P. Chowdhury, and M. A. M. Hasan, “A robust encoder decoder based weighted segmentation and dual staged feature fusion based meta classification for breast cancer utilizing ultrasound imaging,” Intell. Syst. with Appl., vol. 22, no. June, p. 200367, Jun. 2024, doi: 10.1016/j.iswa.2024.200367.
[9] Z. Ji et al., “BGRD-TransUNet: A Novel TransUNet-Based Model for Ultrasound Breast Lesion Segmentation,” IEEE Access, vol. 12, pp. 31182–31196, 2024, doi: 10.1109/ACCESS.2024.3368170.
[10] K. Hu, X. Zhang, D. Lee, D. Xiong, Y. Zhang, and X. Gao, “Boundary-Guided and Region-Aware Network With Global Scale-Adaptive for Accurate Segmentation of Breast Tumors in Ultrasound Images,” IEEE J. Biomed. Heal. Informatics, vol. 27, no. 9, pp. 4421–4432, Sep. 2023, doi: 10.1109/JBHI.2023.3285789.
[11] A. Sulaiman et al., “Attention based UNet model for breast cancer segmentation using BUSI dataset,” Sci. Rep., vol. 14, no. 1, p. 22422, Sep. 2024, doi: 10.1038/s41598-024-72712-5.
[12] G. Chen, L. Li, Y. Dai, J. Zhang, and M. H. Yap, “AAU-Net: An Adaptive Attention U-Net for Breast Lesions Segmentation in Ultrasound Images,” IEEE Trans. Med. Imaging, vol. 42, no. 5, pp. 1289–1300, May 2023, doi: 10.1109/TMI.2022.3226268.
[13] A. A. Hekal, A. Elnakib, H. E.-D. Moustafa, and H. M. Amer, “Breast Cancer Segmentation From Ultrasound Images Using Deep Dual-Decoder Technology With Attention Network,” IEEE Access, vol. 12, pp. 10087–10101, 2024, doi: 10.1109/ACCESS.2024.3351564.
[14] J. Wang, J. Liang, Y. Xiao, J. T. Zhou, Z. Fang, and F. Yang, “TaiChiNet: Negative-Positive Cross-Attention Network for Breast Lesion Segmentation in Ultrasound Images,” IEEE J. Biomed. Heal. Informatics, vol. 28, no. 3, pp. 1516–1527, Mar. 2024, doi: 10.1109/JBHI.2024.3352984.
[15] Z. Ning, S. Zhong, Q. Feng, W. Chen, and Y. Zhang, “SMU-Net: Saliency-Guided Morphology-Aware U-Net for Breast Lesion Segmentation in Ultrasound Image,” IEEE Trans. Med. Imaging, vol. 41, no. 2, pp. 476–490, Feb. 2022, doi: 10.1109/TMI.2021.3116087.
[16] M. J. Umer, M. I. Sharif, and J. Kim, “Breast Cancer Segmentation From Ultrasound Images Using Multiscale Cascaded Convolution With Residual Attention-Based Double Decoder Network,” IEEE Access, vol. 12, pp. 107888–107902, 2024, doi: 10.1109/ACCESS.2024.3429386.
[17] J. Sun et al., “DDRA-Net: Dual-Channel Deep Residual Attention UPerNet for Breast Lesions Segmentation in Ultrasound Images,” IEEE Access, vol. 12, pp. 43691–43703, 2024, doi: 10.1109/ACCESS.2024.3373551.
[18] Q. Qin et al., “MBE-UNet: Multi-Branch Boundary Enhanced U-Net for Ultrasound Segmentation,” IEEE J. Biomed. Heal. Informatics, vol. 30, no. 1, pp. 575–585, Jan. 2026, doi: 10.1109/JBHI.2025.3589293.
[19] S. Abuowaida, H. A. Owida, D. M. Alsekait, N. Alshdaifat, D. S. AbdElminaam, and M. Alshinwan, “UltraSegNet: A Hybrid Deep Learning Framework for Enhanced Breast Cancer Segmentation and Classification on Ultrasound Images,” Comput. Mater. Contin., vol. 83, no. 2, pp. 3303–3333, Apr. 2025, doi: 10.32604/cmc.2025.063470.
[20] J. Huang et al., “UltraMamba: Mamba-Based Multimodal Ultrasound Image Adaptive Fusion for Breast Lesion Segmentation,” IEEE Trans. Med. Imaging, vol. 45, no. 5, pp. 2360–2372, May 2026, doi: 10.1109/TMI.2026.3653779.
[21] H. Wu, X. Huang, X. Guo, Z. Wen, and J. Qin, “Cross-Image Dependency Modeling for Breast Ultrasound Segmentation,” IEEE Trans. Med. Imaging, vol. 42, no. 6, pp. 1619–1631, Jun. 2023, doi: 10.1109/TMI.2022.3233648.
[22] S. Hossain et al., “Automated breast tumor ultrasound image segmentation with hybrid UNet and classification using fine-tuned CNN model,” Heliyon, vol. 9, no. 11, p. e21369, Nov. 2023, doi: 10.1016/j.heliyon.2023.e21369.
[23] M. R. Islam et al., “Enhancing breast cancer segmentation and classification: An Ensemble Deep Convolutional Neural Network and U-net approach on ultrasound images,” Mach. Learn. with Appl., vol. 16, no. June, p. 100555, Jun. 2024, doi: 10.1016/j.mlwa.2024.100555.
[24] A. Su et al., “Multi-task learning for multi-scale breast cancer ultrasound image segmentation and classification based on visual perception,” Biomed. Signal Process. Control, vol. 110, no. December, p. 108212, Dec. 2025, doi: 10.1016/j.bspc.2025.108212.
[25] Z. Zhuang, Z. Yang, A. N. J. Raj, C. Wei, P. Jin, and S. Zhuang, “Breast ultrasound tumor image classification using image decomposition and fusion based on adaptive multi-model spatial feature fusion,” Comput. Methods Programs Biomed., vol. 208, no. September, p. 106221, Sep. 2021, doi: 10.1016/j.cmpb.2021.106221.
[26] F. Taheri and K. Rahbar, “Improving breast cancer classification in fine-grain ultrasound images through feature discrimination and a transfer learning approach,” Biomed. Signal Process. Control, vol. 106, no. August, p. 107690, Aug. 2025, doi: 10.1016/j.bspc.2025.107690.
[27] P. Zhang, Y. Dong, J. Li, L. Jiang, M. Hu, and Y. Ping, “MSSM-MFP: Medical semantic segmentation model based on multiscale fusion perception,” Biomed. Signal Process. Control, vol. 112, no. February, p. 108481, Feb. 2026, doi: 10.1016/J.BSPC.2025.108481.
[28] L. Wu et al., “Breast tumor detection in ultrasound images with anatomical prior knowledge,” Image Vis. Comput., vol. 162, no. October, p. 105724, Oct. 2025, doi: 10.1016/J.IMAVIS.2025.105724.
[29] Y. Wang and Y. Yao, “Breast lesion detection using an anchor-free network from ultrasound images with segmentation-based enhancement,” Sci. Rep., vol. 12, no. 1, p. 14720, Aug. 2022, doi: 10.1038/s41598-022-18747-y.
[30] S. Jiang, K. Gan, Q. Wang, L. Gong, W. Kong, and J. Yuan, “Advancing breast nodule detection in ultrasound images with multi-scale feature extraction,” Biomed. Signal Process. Control, vol. 110, no. December, p. 108196, Dec. 2025, doi: 10.1016/j.bspc.2025.108196.
[31] N. Ravi et al., “SAM 2: Segment Anything in Images and Videos,” in 13th International Conference on Learning Representations, ICLR 2025, International Conference on Learning Representations, ICLR, Oct. 2024, pp. 41175–41218. Accessed: Jun. 19, 2026. [Online]. Available: http://arxiv.org/abs/2408.00714
[32] J. Zhu, A. Hamdi, Y. Qi, Y. Jin, and J. Wu, “Medical SAM 2: Segment medical images as video via Segment Anything Model 2,” pp. 1–15, Dec. 2024, Accessed: Jun. 19, 2026. [Online]. Available: http://arxiv.org/abs/2408.00874
[33] J. Ma et al., “MedSAM2: Segment Anything in 3D Medical Images and Videos,” pp. 1–13, Apr. 2025, Accessed: Jun. 19, 2026. [Online]. Available: https://arxiv.org/pdf/2504.03600
[34] E. Chukwujindu, K. Faiz, A. De Sequeira, S. Chidom, and H. Faiz, “Improving medical image segmentation with SAM2: analyzing the impact of object characteristics and finetuning on multi-planar datasets.,” Eur. J. Radiol. Artif. Intell., vol. 3, no. September, p. 100034, Sep. 2025, doi: 10.1016/j.ejrai.2025.100034.
[35] Y. Zhou et al., “Semiautomated segmentation of breast tumor on automatic breast ultrasound image using a large-scale model with customized modules,” Sci. Rep., vol. 15, no. 1, p. 17329, May 2025, doi: 10.1038/s41598-025-97098-w.
[36] J. Wu et al., “Medical SAM adapter: Adapting segment anything model for medical image segmentation,” Med. Image Anal., vol. 102, no. May, p. 103547, May 2025, doi: 10.1016/j.media.2025.103547.
[37] Z. Zhong, Z. Tang, T. He, H. Fang, and C. Yuan, “Convolution Meets LoRA: Parameter Efficient Finetuning for Segment Anything Model,” in 12th International Conference on Learning Representations, ICLR 2024, International Conference on Learning Representations, ICLR, Jan. 2024, pp. 1–25. Accessed: Jun. 19, 2026. [Online]. Available: https://arxiv.org/pdf/2401.17868
[38] S. N. Gowda and D. A. Clifton, “CC-SAM: SAM with Cross-Feature Attention and Context for Ultrasound Image Segmentation,” in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 15103 LNCS, Springer Science and Business Media Deutschland GmbH, 2025, pp. 108–124. doi: 10.1007/978-3-031-72995-9_7.
[39] Z. Tu, L. Gu, X. Wang, and B. Jiang, “Ultrasound SAM Adapter: Adapting SAM for Breast Lesion Segmentation in Ultrasound Images,” pp. 1–9, Apr. 2024, Accessed: Jun. 19, 2026. [Online]. Available: http://arxiv.org/abs/2404.14837
[40] T. Jiang et al., “A segmentation knowledge-based global-local attention network for tumor classification in breast ultrasound images,” Pattern Recognit., vol. 171, no. March, p. 112152, Mar. 2026, doi: 10.1016/j.patcog.2025.112152.
[41] L. Guo, H. Zhang, and C. Ma, “ESAM2-BLS: Enhanced segment anything model 2 for efficient breast lesion segmentation in ultrasound imaging,” Comput. Med. Imaging Graph., vol. 126, no. December, p. 102654, Dec. 2025, doi: 10.1016/j.compmedimag.2025.102654.
[42] X. Yue et al., “Morphology-enhanced CAM-guided SAM for weakly supervised breast lesion segmentation,” Biomed. Signal Process. Control, vol. 116, no. May, p. 109509, May 2026, doi: 10.1016/j.bspc.2026.109509.
[43] X. Liu, M. Hussain, J. Huang, Q. Li, M. T. Khan, and H. Wu, “Auto-BUSAM: Auto-segmentation of attention-diverted low-contrast breast ultrasound images,” Displays, vol. 92, no. April, p. 103314, Apr. 2026, doi: 10.1016/j.displa.2025.103314.
[44] B. Liu, H. Chen, T. Zhu, Z. Ye, H. Cui, and K. Wang, “Yolo-HLSAM: Adapting foundation segment anything model for semi-automatic detection and segmentation of breast cancer microcalcification clusters,” Biomed. Signal Process. Control, vol. 111, no. January, p. 108300, Jan. 2026, doi: 10.1016/j.bspc.2025.108300.
[45] D. Yin, Q. Zheng, L. Chen, Y. Hu, and Q. Wang, “APG-SAM: Automatic prompt generation for SAM-based breast lesion segmentation with boundary-aware optimization,” Expert Syst. Appl., vol. 276, no. June, p. 127048, Jun. 2025, doi: 10.1016/j.eswa.2025.127048.
[46] W. Al-Dhabyani, M. Gomaa, H. Khaled, and A. Fahmy, “Dataset of breast ultrasound images,” Data Br., vol. 28, no. February, p. 104863, Feb. 2020, doi: 10.1016/j.dib.2019.104863.
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
___________________________________________________________
International Journal of Advances in Intelligent Informatics
ISSN 2442-6571 (print) | 2548-3161 (online)
Organized by UAD and ASCEE Computer Society
Published by Universitas Ahmad Dahlan
W: http://ijain.org
E: info@ijain.org (paper handling issues)
andri.pranolo.id@ieee.org (publication issues)
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0