Abstract
Purpose
Recently, machine learning has outperformed established tools for automated segmentation in medical imaging. However, segmentation of cardiac chambers still proves challenging due to the variety of contrast agent injection protocols used in clinical practice, inducing disparities of contrast between cavities. Hence, training a generalist network requires large training datasets representative of these protocols. Furthermore, segmentation on unenhanced CT scans is further hindered by the challenge of obtaining ground truths from these images. Newly available spectral CT scanners allow innovative image reconstructions such as virtual non-contrast (VNC) imaging, mimicking non-contrasted conventional CT studies from a contrasted scan. Recent publications have demonstrated that networks can be trained using VNC to segment contrasted and unenhanced conventional CT scans to reduce annotated data requirements and the need for annotations on unenhanced scans. We propose an extensive evaluation of this statement.
Method
We undertake multiple trainings of a 3D multi-label heart segmentation network with (HU-VNC) and without (HUonly) VNC as augmentation, using decreasing training dataset sizes (114, 76, 57, 38, 29, 19 patients). At each step, both networks are tested on a multi-vendor, multi-centric dataset of 122 patients, including different protocols: pulmonary embolism (PE), chest-abdomen-pelvis (CAP), heart CT angiography (CTA) and true non-contrast scans (TNC). An in-depth comparison of resulting Dice coefficients and distance metrics is performed for the networks trained on the largest dataset.
Results
HU-VNC-trained on 57 patients significantly outperforms HUonly trained on 114 regarding CAP and TNC scans (mean Dice coefficients of 0.881/0.835 and 0.882/0.416, respectively). When trained on the largest dataset, significant improvements in all labels are noted for TNC and CAP scans (mean Dice coefficient of 0.882/0.416 and 0.891/0.835, respectively).
Conclusion
Adding VNC images as training augmentation allows the network to perform on unenhanced scans and improves segmentations on other imaging protocols, while using a reduced training dataset.
Similar content being viewed by others
Code availability
Philips IntelliSpace Portal 11.1 and 3D-Slicer 4.10.2 were used to create segmentation ground truths.
References
Chen C, Qin C, Qiu H, Tarroni G, Duan J, Bai W, Rueckert D (2020) Deep learning for cardiac image segmentation: a review. Front Cardiovasc Med 7:25. https://doi.org/10.3389/fcvm.2020.00025
Zhuang X, Li L, Payer C, Štern D, Urschler M, Heinrich MP, Oster J, Wang C, Smedby Ö, Bian C, Yang X, Heng P-A, Mortazi A, Bagci U, Yang G, Sun C, Galisot G, Ramel J-Y, Brouard T, Tong Q, Si W, Liao X, Zeng G, Shi Z, Zheng G, Wang C, MacGillivray T, Newby D, Rhode K, Ourselin S, Mohiaddin R, Keegan J, Firmin D, Yang G (2019) Evaluation of algorithms for multi-modality whole heart segmentation: an open-access grand challenge. Med Image Anal 58:101537. https://doi.org/10.1016/j.media.2019.101537
Singh G, Al’Aref SJ, Van Assen M, Kim TS, van Rosendael A, Kolli KK, Dwivedi A, Maliakal G, Pandey M, Wang J, Do V, Gummalla M, De Cecco CN, Min JK (2018) Machine learning in cardiac CT: basic concepts and contemporary data. J Cardiovasc Comput Tomogr 12:192–201. https://doi.org/10.1016/j.jcct.2018.04.010
Shameer K, Johnson KW, Glicksberg BS, Dudley JT, Sengupta PP (2018) Machine learning in cardiovascular medicine: are we there yet? Heart 104:1156–1164. https://doi.org/10.1136/heartjnl-2017-311198
Kurkure U, Avila-Montes OC, Kakadiaris IA (2008) Automated segmentation of thoracic aorta in non-contrast CT images. In: 2008 5th IEEE International Symposium on Biomedical Imaging: From Nano to Macro. pp 29–32
Kurugol S, San Jose Estepar R, Ross J, Washko GR (2012) Aorta segmentation with a 3D level set approach and quantification of aortic calcifications in non-contrast chest CT. In: 2012 Annual International Conference of the IEEE Engineering in Medicine and Biology Society, pp. 2343–2346
Xie Y, Padgett J, Biancardi AM, Reeves AP (2014) Automated aorta segmentation in low-dose chest CT images. Int J Comput Assist Radiol Surg 9:211–219. https://doi.org/10.1007/s11548-013-0924-5
Sedghi Gamechi Z, Bons LR, Giordano M, Bos D, Budde RPJ, Kofoed KF, Pedersen JH, Roos-Hesselink JW, de Bruijne M (2019) Automated 3D segmentation and diameter measurement of the thoracic aorta on non-contrast enhanced CT. Eur Radiol 29:4613–4623. https://doi.org/10.1007/s00330-018-5931-z
Wang C, Smedby Ö (2014) Automatic multi-organ segmentation in non-enhanced ct datasets using hierarchical shape priors. In: 2014 22nd international conference on pattern recognition, pp. 3327–3332
Dey D, Ramesh A, Slomka PJ, Nakazato R, Cheng VY, Germano G, Berman DS (2010) Automated algorithm for atlas-based segmentation of the heart and pericardium from non-contrast CT. In: Proc. SPIE—Int. Soc. Opt. Eng., vol. 7623, p. 762337. https://doi.org/10.1117/12.844810
Luo Y, Xu Y, Liao Z, Gomez D, Wang J, Jiang W, Zhou R, Williamson R, Court LE, Yang J (2019) Automatic segmentation of cardiac substructures from noncontrast CT images: accurate enough for dosimetric analysis? Acta Oncol Stockh Swed 58:81–87. https://doi.org/10.1080/0284186X.2018.1521985
Shahzad R, Bos D, Budde RPJ, Pellikaan K, Niessen WJ, van der Lugt A, van Walsum T (2017) Automatic segmentation and quantification of the cardiac structures from non-contrast-enhanced cardiac CT scans. Phys Med Biol 62:3798–3813. https://doi.org/10.1088/1361-6560/aa63cb
Chen S, Zhong X, Hu S, Dorn S, Kachelrieß M, Lell M, Maier A (2020) Automatic multi-organ segmentation in dual-energy CT (DECT) with dedicated 3D fully convolutional DECT networks. Med Phys 47:552–562. https://doi.org/10.1002/mp.13950
González Sánchez JC, Magnusson M, Sandborg M, Carlsson Tedgren Å, Malusek A (2020) Segmentation of bones in medical dual-energy computed tomography volumes using the 3D U-Net. Phys Med 69:241–247. https://doi.org/10.1016/j.ejmp.2019.12.014
Lartaud P-J, Rouchaud A, Rouet J-M, Nempont O, Boussel L (2019) Spectral CT based training dataset generation and augmentation for conventional CT vascular segmentation. In: Shen D, Liu T, Peters TM, Staib LH, Essert C, Zhou S, Yap P-T, Khan A (eds) Medical image computing and computer assisted intervention—MICCAI 2019. Springer, Cham, pp 768–775
Bruns S, Wolterink JM, van Hamersvelt RW, Leiner T, Išgum I (2019) CNN-based segmentation of the cardiac chambers and great vessels in non-contrast-enhanced cardiac CT. In: International conference on medical imaging with deep learning—extended abstract track, London, United Kingdom
Bruns S, Wolterink JM, Takx RAP, van Hamersvelt RW, Suchá D, Viergever MA, Leiner T, Išgum I (2020) Deep learning from dual-energy information for whole-heart segmentation in dual-energy and single-energy non-contrast-enhanced cardiac CT. Med Phys 47:5048–5060. https://doi.org/10.1002/mp.14451
McCollough CH, Leng S, Yu L, Fletcher JG (2015) Dual- and multi-energy CT: principles, technical approaches, and clinical applications. Radiology 276:637–653. https://doi.org/10.1148/radiol.2015142631
Alvarez RE, Macovski A (1976) Energy-selective reconstructions in X-ray computerized tomography. Phys Med Biol 21:733–744. https://doi.org/10.1088/0031-9155/21/5/002
Si-Mohamed S, Dupuis N, Tatard-Leitman V, Rotzinger D, Boccalini S, Dion M, Vlassenbroek A, Coulon P, Yagil Y, Shapira N, Douek P, Boussel L (2019) Virtual versus true non-contrast dual-energy CT imaging for the diagnosis of aortic intramural hematoma. Eur Radiol 29:6762–6771. https://doi.org/10.1007/s00330-019-06322-5
Sauter AP, Muenzel D, Dangelmaier J, Braren R, Pfeiffer F, Rummeny EJ, Noël PB, Fingerle AA (2018) Dual-layer spectral computed tomography: virtual non-contrast in comparison to true non-contrast images. Eur J Radiol 104:108–114. https://doi.org/10.1016/j.ejrad.2018.05.007
Wang C, MacGillivray T, Macnaught G, Yang G, Newby D (2018) A two-stage 3D Unet framework for multi-class segmentation on full resolution image. arXiv:1804.04341 Cs
Xu Z, Wu Z, Feng J (2018) CFUN: combining faster R-CNN and U-net network for efficient whole heart segmentation. arXiv:1812.04914 Cs
Ye C, Wang W, Zhang S, Wang K (2019) Multi-depth fusion network for whole-heart CT image segmentation. IEEE Access 7:23421–23429. https://doi.org/10.1109/ACCESS.2019.2899635
Yang X, Bian C, Yu L, Ni D, Heng P-A (2018) 3D convolutional networks for fully automatic fine-grained whole heart partition. In: Pop M, Sermesant M, Jodoin P-M, Lalande A, Zhuang X, Yang G, Young A, Bernard O (eds) Statistical atlases and computational models of the heart. ACDC and MMWHS challenges. Springer, Cham, pp 181–189
He K, Sun J, Tang X (2013) Guided image filtering. IEEE Trans Pattern Anal Mach Intell 35:1397–1409. https://doi.org/10.1109/TPAMI.2012.213
Taha AA, Hanbury A (2015) Metrics for evaluating 3D medical image segmentation: analysis, selection, and tool. BMC Med Imaging 15:29. https://doi.org/10.1186/s12880-015-0068-x
Funding
This work was partially funded by a grant from the Association Nationale de la Recherche et de la Technologie (#2018/1716).
Author information
Authors and Affiliations
Contributions
Authors D. Hallé and A. Schleef contributed equally to this work.
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Consent to participate
Informed consent was obtained from all individual participants included in the study, following MR004 protocol.
Consent for publication
Informed consent was obtained from all individual participants included in the study, following MR004 protocol.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Lartaud, PJ., Hallé, D., Schleef, A. et al. Spectral augmentation for heart chambers segmentation on conventional contrasted and unenhanced CT scans: an in-depth study. Int J CARS 16, 1699–1709 (2021). https://doi.org/10.1007/s11548-021-02468-0
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11548-021-02468-0