Abstract
Ultrasound Confidence Map (CM) is an image representation that indicates the reliability of the pixel intensity values presented within ultrasound B-mode images. Those maps are highly correlated with the probability of sound reaching specific depths. Commonly, CM is calculated based only on the B-mode images, without anatomy awareness. Without clear anatomical landmarks or contextual information, CMs might misrepresent the certainty of features detected within the ultrasound images. We propose a novel deep-learning approach for CM calculation that is specific to the anatomy and based on physical principles of echo propagation. We rely on the physics-inspired intermediate representation maps of Ultra-Nerf to compute CMs with observation-angle awareness, similar to the clinical practice. Our method outperforms other methods on downstream tasks such as shadow segmentation and compounding. Additionally, we open-source the code and a tracked ultrasound dataset to promote more research in this direction at https://github.com/MrGranddy/Redefining-Confidence-Maps.
B. Yesilkaynak and V. G. Duque—Both authors share first authorship.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
Notes
- 1.
Anisotropy refers to the variance in visual characteristics like textures and edges depending on the direction observed.
References
Patil, P., Dasgupta, B.: Role of diagnostic ultrasound in the assessment of musculoskeletal diseases. Therap. Adv. Musculoskeletal Disease 4(5), 341–355 (2012)
Jang, J.K., Kim, S.Y., Yoo, I.W., Cho, Y.B., Kang, H.J., Lee, D.H.: Diagnostic performance of ultrasound attenuation imaging for assessing low-grade hepatic steatosis. Eur. Radiol. 1–8 (2022)
Wei-Ting, W., Chang, K.-V., Hsu, Y.-C., Hsu, P.-C., Ricci, V., Özçakar, L.: Artifacts in musculoskeletal ultrasonography: from physics to clinics. Diagnostics 10(9), 645 (2020)
Hellier, P., Coupé, P., Meyer, P., Morandi, X., Collins, D.L.: Acoustic shadows detection, application to accurate reconstruction of 3D intraoperative ultrasound. In: 2008 5th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, pp. 1569–1572. IEEE (2008)
Berge, C.S., Kapoor, A., Navab, N.: Orientation-driven ultrasound compounding using uncertainty information. In: Stoyanov, D., Collins, D.L., Sakuma, I., Abolmaesumi, P., Jannin, P. (eds.) IPCAI 2014. LNCS, vol. 8498, pp. 236–245. Springer, Cham (2014). https://doi.org/10.1007/978-3-319-07521-1_25
Wein, W., Karamalis, A., Baumgartner, A., Navab, N.: Automatic bone detection and soft tissue aware ultrasound-CT registration for computer-aided orthopedic surgery. Int. J. Comput. Assist. Radiol. Surg. 10, 971–979 (2015)
Klein, T., Wells, W.M.: RF ultrasound distribution-based confidence maps. In: Navab, N., Hornegger, J., Wells, W.M., Frangi, A.F. (eds.) MICCAI 2015, Part II 18. LNCS, vol. 9350, pp. 595–602. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-24571-3_71
Duque, V.G., Zirus, L., Velikova, Y., Navab, N., Mateus, D.: Can ultrasound confidence maps predict sonographers’ labeling variability? In: Kainz, B., Noble, A., Schnabel, J., Khanal, B., Müller, J.P., Day, T. (eds.) ASMUS 2023. LNCS, vol. 14337, pp. 175–184. Springer, Cham (2023). https://doi.org/10.1007/978-3-031-44521-7_17
Chatelain, P., Krupa, A., Navab, N.: Confidence-driven control of an ultrasound probe. IEEE Trans. Rob. 33(6), 1410–1424 (2017)
Virga, S., et al.: Automatic force-compliant robotic ultrasound screening of abdominal aortic aneurysms. In: 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 508–513. IEEE (2016)
Meng, Q., et al.: Weakly supervised estimation of shadow confidence maps in fetal ultrasound imaging. IEEE Trans. Med. Imaging (2019)
Karamalis, A., Wein, W., Klein, T., Navab, N.: Ultrasound confidence maps using random walks. Med. Image Anal. 16(6), 1101–1112 (2012)
Hung, A.L.Y., Chen, W., Galeotti, J.: Ultrasound confidence maps of intensity and structure based on directed acyclic graphs and artifact models. In: 2021 IEEE 18th International Symposium on Biomedical Imaging (ISBI), pp. 697–701. IEEE (2021)
Sitzmann, V., Martel, J., Bergman, A., Lindell, D., Wetzstein, G.: Implicit neural representations with periodic activation functions. In: Advances in Neural Information Processing Systems, vol. 33, pp. 7462–7473 (2020)
Yeung, P.-H., et al.: Implicitvol: sensorless 3D ultrasound reconstruction with deep implicit representation. arXiv preprint arXiv:2109.12108 (2021)
Alblas, D., Brune, C., Yeung, K.K., Wolterink, J.M.: Going off-grid: continuous implicit neural representations for 3d vascular modeling. In: Camara, O., et al. (eds.) STACOM 2022. LNCS, vol. 13593, pp. 79–90. Springer, Cham (2022). https://doi.org/10.1007/978-3-031-23443-9_8
Velikova, Y., Azampour, M.F., Simson, W., Duque, V.G., Navab, N.: Lotus: learning to optimize task-based us representations. In: Greenspan, H., et al. (eds.) MICCAI 2023. LNCS, vol. 14220, pp. 435–445. Springer, Cham (2023). https://doi.org/10.1007/978-3-031-43907-0_42
Wysocki, M., Azampour, M.F., Eilers, C., Busam, B., Salehi, M., Navab, N.: Ultra-nerf: neural radiance fields for ultrasound imaging. In: Medical Imaging with Deep Learning, pp. 382–401. PMLR (2024)
Alsinan, A.Z., Patel, V.M., Hacihaliloglu, I.: Bone shadow segmentation from ultrasound data for orthopedic surgery using GAN. Int. J. Comput. Assist. Radiol. Surg. 15(9), 1477–1485 (2020)
Berton, F., Cheriet, F., Miron, M.-C., Laporte, C.: Segmentation of the spinous process and its acoustic shadow in vertebral ultrasound images. Comput. Biol. Med. 72, 201–211 (2016)
Wysocki, M.: Neural radiance fields for ultrasound imaging. Master’s thesis, Technische Universität München (2023)
Reinke, A., et al.: Common limitations of performance metrics in biomedical image analysis. In: Medical Imaging with Deep Learning (2021)
Quader, N., Hodgson, A., Abugharbieh, R.: Confidence weighted local phase features for robust bone surface segmentation in ultrasound. In: Linguraru, M.G., et al. (eds.) CLIP 2014. LNCS, vol. 8680, pp. 76–83. Springer, Cham (2014). https://doi.org/10.1007/978-3-319-13909-8_10
Acknowledgments
This work supported by the European Regional Development. Fund, the Pays de la Loire region on the Connect Talent scheme (MILCOM Project) and Nantes Métropole (Convention 2017-10470). Additionally, this work is partially supported by the Bavarian Ministry of Economics, State Development, and Energy project HINAV (LSM-2303-0005).
Images were rendered using ImFusion software, version 2.36.3, from ImFusion GmbH, Munich, Germany [6].
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2024 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Yesilkaynak, V.B., Duque, V.G., Wysocki, M., Velikova, Y., Mateus, D., Navab, N. (2024). Ultrasound Confidence Maps with Neural Implicit Representation. In: Yap, M.H., Kendrick, C., Behera, A., Cootes, T., Zwiggelaar, R. (eds) Medical Image Understanding and Analysis. MIUA 2024. Lecture Notes in Computer Science, vol 14860. Springer, Cham. https://doi.org/10.1007/978-3-031-66958-3_7
Download citation
DOI: https://doi.org/10.1007/978-3-031-66958-3_7
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-66957-6
Online ISBN: 978-3-031-66958-3
eBook Packages: Computer ScienceComputer Science (R0)