Abstract
Mitral regurgitation (MR) is a frequent indication for valve surgery. One of its treatments, mitral valve performed with endoscopic video recordings, is a complex minimally invasive procedure which is facing the problem of data availability and data privacy. Therefore, the simulation cases are widely used to form surgery training and planning. However, the cross-domain gap may affect the performance significantly as Deep Learning methods rely heavily on data. We propose to develop an algorithm to reduce the domain gap between simulation and intra-operative cases. The task is to learn the distance and location information of the points and predict a series of 2D landmarks’ location, the coordinates of the landmark were both marked on real and simulate dataset by the AdaptOR Challenge organizer. Our work has merged the data from both domains by using a relation heatmap generation algorithm, which can generate a relation key point heatmap based on the distance measurement of landmarks and explicitly represent the geometric relation between landmarks. The MSE loss function is used to minimize the error between the ground-truth and predicted heatmaps. We test our methods on a challenge dataset, in which the model has achieved a good F1 score of 66.19%.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Ancona, R., Pinto, S.C.: Mitral valve incompetence: epidemiology and causes. https://www.escardio.org/Journals/E-Journal-of-Cardiology-Practice/Volume-16/Mitral-valve-incompetence-epidemiology-and-causes (2018)
Chen, L.C., Zhu, Y., Papandreou, G., Schroff, F., Adam, H.: Encoder-decoder with atrous separable convolution for semantic image segmentation. In: Proceedings of the European Conference on Computer Vision (ECCV), pp. 801–818 (2018)
Engelhardt, S., et al.: Deep Generative Model Challenge for Domain Adaptation in Surgery 2021 (March 2021). https://doi.org/10.5281/zenodo.4646979
Engelhardt, S., Sauerzapf, S., Brčić, A., Karck, M., Wolf, I., De Simone, R.: Replicated mitral valve models from real patients offer training opportunities for minimally invasive mitral valve repair. Interact. Cardiovasc. Thorac. Surg. 29(1), 43–50 (2019)
Goodfellow, I., et al.: Generative adversarial nets. Adv. Neural Inf. Process. Syst. 27 (2014)
He, J., Jia, X., Chen, S., Liu, J.: Multi-source domain adaptation with collaborative learning for semantic segmentation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 11008–11017 (2021)
Law, H., Deng, J.: Cornernet: detecting objects as paired keypoints. In: Proceedings of the European Conference on Computer Vision (ECCV), pp. 734–750 (2018)
Sun, K., Xiao, B., Liu, D., Wang, J.: Deep high-resolution representation learning for human pose estimation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 5693–5703 (2019)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Wang, J., Wang, H., Mu, R., Wang, L. (2021). Cross-Domain Landmarks Detection in Mitral Regurgitation. In: Engelhardt, S., et al. Deep Generative Models, and Data Augmentation, Labelling, and Imperfections. DGM4MICCAI DALI 2021 2021. Lecture Notes in Computer Science(), vol 13003. Springer, Cham. https://doi.org/10.1007/978-3-030-88210-5_12
Download citation
DOI: https://doi.org/10.1007/978-3-030-88210-5_12
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-88209-9
Online ISBN: 978-3-030-88210-5
eBook Packages: Computer ScienceComputer Science (R0)