{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/deep-learning-cardiac-motion-analysis-for","title":"Deep learning cardiac motion analysis for human survival prediction","arxiv_id":"1810.03382","date":"2018-10-08","proceeding":null,"authors":["Ghalib A. Bello","Timothy J. W. Dawes","Jinming Duan","Carlo Biffi","Antonio de Marvao","Luke S. G. E. Howard","J. Simon R. Gibbs","Martin R. Wilkins","Stuart A. Cook","Daniel Rueckert","Declan P. O'Regan"],"abstract":"Motion analysis is used in computer vision to understand the behaviour of\nmoving objects in sequences of images. Optimising the interpretation of dynamic\nbiological systems requires accurate and precise motion tracking as well as\nefficient representations of high-dimensional motion trajectories so that these\ncan be used for prediction tasks. Here we use image sequences of the heart,\nacquired using cardiac magnetic resonance imaging, to create time-resolved\nthree-dimensional segmentations using a fully convolutional network trained on\nanatomical shape priors. This dense motion model formed the input to a\nsupervised denoising autoencoder (4Dsurvival), which is a hybrid network\nconsisting of an autoencoder that learns a task-specific latent code\nrepresentation trained on observed outcome data, yielding a latent\nrepresentation optimised for survival prediction. To handle right-censored\nsurvival outcomes, our network used a Cox partial likelihood loss function. In\na study of 302 patients the predictive accuracy (quantified by Harrell's\nC-index) was significantly higher (p < .0001) for our model C=0.73 (95$\\%$ CI:\n0.68 - 0.78) than the human benchmark of C=0.59 (95$\\%$ CI: 0.53 - 0.65). This\nwork demonstrates how a complex computer vision task using high-dimensional\nmedical image data can efficiently predict human survival.","url_abs":"http://arxiv.org/abs/1810.03382v1","url_pdf":"http://arxiv.org/pdf/1810.03382v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"deep-learning-cardiac-motion-analysis-for","repo_url":"https://github.com/UK-Digital-Heart-Project/4Dsurvival","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"GPL-3.0"}}],"tasks":[{"task_slug":"deep-learning","task_name":"Deep Learning"},{"task_slug":"denoising","task_name":"Denoising"},{"task_slug":"survival-prediction","task_name":"Survival Prediction"}],"methods":[{"method_slug":"denoising-autoencoder","method_name":"Denoising Autoencoder"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1810.03382","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}