Abstract
3D ultrasound systems have been widely used for fetal brain structures’ analysis. However, the obtained images present several artifacts such as multiplicative noise and acoustic shadows appearing as a function of acquisition angle. The purpose of this research is to merge several partially occluded ultrasound volumes, acquired by placing the transducer at different projections of the fetal head, to compound a new US volume containing the whole brain anatomy. To achieve this, the proposed methodology consists on the pipeline of four convolutional neural networks (CNN). Two CNNs are used to carry out fetal skull segmentations, by incorporating an incidence angle map and the segmented structures are then described with a Gaussian mixture model (GMM). For multiple US volumes registration, a feature set, based on distance maps computed from the GMM centroids is proposed. The third CNN learns the relation between distance maps of the volumes to be registered and estimates optimal rotation and translation parameters. Finally, the weighted root mean square is proposed as composition operator and weighting factors are estimated with the last CNN, which assigns a higher weight to those regions containing brain tissue and less ponderation to acoustic shadowed areas. The procedure was qualitatively and quantitatively validated in a set of fetal volumes obtained during gestation’s second trimester. Results show registration errors of 1.31 ± 0.2 mm and an increase of image sharpness of 34.9% compared to a single acquisition and of 25.2% compared to root mean square compounding.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsReferences
Besl, P., McKay, H.: A method for registration of 3-D shapes. IEEE Trans. Pattern Anal. Mach. Intell. 14(2), 239–256 (1992)
Cen, F., Jiang, Y., Zhang, Z., Tsui, H.T., Lau, T.K., Xie, H.: Robust registration of 3-D ultrasound images based on Gabor filter and mean-shift method. In: Sonka, M., Kakadiaris, I.A., Kybic, J. (eds.) CVAMIA/MMBIA - 2004. LNCS, vol. 3117, pp. 304–316. Springer, Heidelberg (2004). https://doi.org/10.1007/978-3-540-27816-0_26
Cen, F., Jiang, Y., Zhang, Z., Tsui, H.T.: Shape and pixel-property based automatic affine registration between ultrasound images of different fetal head. In: Yang, G.-Z., Jiang, T.-Z. (eds.) MIAR 2004. LNCS, vol. 3150, pp. 261–269. Springer, Heidelberg (2004). https://doi.org/10.1007/978-3-540-28626-4_32
Cerrolaza, J.J., et al.: Deep learning with ultrasound physics for fetal skull segmentation. In: 2018 IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018), pp. 564–567, April 2018. https://doi.org/10.1109/ISBI.2018.8363639
Chen, H.C., et al.: Registration-based segmentation of three-dimensional ultrasound images for quantitative measurement of fetal craniofacial structure. Ultrasound Med. Biol. 38(5), 811–823 (2012)
Contreras Ortiz, S.H., Chiu, T., Fox, M.D.: Ultrasound image enhancement: a review. Biomed. Signal Process. Control 7(5), 419–428 (2012)
Gooding, M.J., Rajpoot, K., Mitchell, S., Chamberlain, P., Kennedy, S.H., Noble, J.A.: Investigation into the fusion of multiple 4-D fetal echocardiography images to improve image quality. Ultrasound Med. Biol. 36(6), 957–966 (2010)
Modat, M., et al.: Global image registration using a symmetric block-matching approach. J. Med. Imaging 1(2), 1–6 (2014)
Perez-Gonzalez, J., Arámbula Cosío, F., Huegel, J., Medina-Bañuelos, V.: Probabilistic learning coherent point drift for 3D ultrasound fetal head registration. Comput. Math. Methods Med. 2020 (2020). https://doi.org/10.1155/2020/4271519
Perez-Gonzalez, J.L., Arámbula Cosío, F., Medina-Bañuelos, V.: Spatial composition of US images using probabilistic weighted means. In: 11th International Symposium on Medical Information Processing and Analysis, International Society for Optics and Photonics, SPIE, vol. 9681, pp. 288–294 (2015). https://doi.org/10.1117/12.2207958
Perez-Gonzalez, J., et al.: Spatial compounding of 3-D fetal brain ultrasound using probabilistic maps. Ultrasound Med. Biol. 44(1), 278–291 (2018). https://doi.org/10.1016/j.ultrasmedbio.2017.09.001
Rajpoot, K., Grau, V., Noble, J.A., Szmigielski, C., Becher, H.: Multiview fusion 3-d echocardiography: improving the information and quality of real-time 3-D echocardiography. Ultrasound Med. Biol. 37(7), 1056–1072 (2011)
Rohling, R., Gee, A., Berman, L.: Three-dimensional spatial compounding of ultrasound images. Med. Image Anal. 1(3), 177–193 (1997)
Wang, Y., Solomon, J.M.: Deep closest point: learning representations for point cloud registration. In: IEEE 15th International Symposium on Biomedical Imaging (ISBI 2018), pp. 564–567, April 2019. https://doi.org/10.1109/ISBI.2018.8363639
Wilhjelm, J., Jensen, M., Jespersen, S., Sahl, B., Falk, E.: Visual and quantitative evaluation of selected image combination schemes in ultrasound spatial compound scanning. IEEE Trans. Med. Imaging 23(2), 181–190 (2004)
Wright, R., et al.: Complete fetal head compounding from multi-view 3D ultrasound. In: Shen, D., et al. (eds.) MICCAI 2019. LNCS, vol. 11766, pp. 384–392. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-32248-9_43
Acknowledgment
This work was supported by UNAM–PAPIIT IA102920 and IT100220. The authors also acknowledge the National Institute of Perinatology of Mexico (INPer) for sharing the Ultrasound images.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Perez-Gonzalez, J., Hevia Montiel, N., Bañuelos, V.M. (2020). Deep Learning Spatial Compounding from Multiple Fetal Head Ultrasound Acquisitions. In: Hu, Y., et al. Medical Ultrasound, and Preterm, Perinatal and Paediatric Image Analysis. ASMUS PIPPI 2020 2020. Lecture Notes in Computer Science(), vol 12437. Springer, Cham. https://doi.org/10.1007/978-3-030-60334-2_30
Download citation
DOI: https://doi.org/10.1007/978-3-030-60334-2_30
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-60333-5
Online ISBN: 978-3-030-60334-2
eBook Packages: Computer ScienceComputer Science (R0)