{"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/fast-multiple-landmark-localisation-using-a","title":"Fast Multiple Landmark Localisation Using a Patch-based Iterative Network","arxiv_id":"1806.06987","date":"2018-06-18","proceeding":null,"authors":["Yuanwei Li","Amir Alansary","Juan J. Cerrolaza","Bishesh Khanal","Matthew Sinclair","Jacqueline Matthew","Chandni Gupta","Caroline Knight","Bernhard Kainz","Daniel Rueckert"],"abstract":"We propose a new Patch-based Iterative Network (PIN) for fast and accurate\nlandmark localisation in 3D medical volumes. PIN utilises a Convolutional\nNeural Network (CNN) to learn the spatial relationship between an image patch\nand anatomical landmark positions. During inference, patches are repeatedly\npassed to the CNN until the estimated landmark position converges to the true\nlandmark location. PIN is computationally efficient since the inference stage\nonly selectively samples a small number of patches in an iterative fashion\nrather than a dense sampling at every location in the volume. Our approach\nadopts a multi-task learning framework that combines regression and\nclassification to improve localisation accuracy. We extend PIN to localise\nmultiple landmarks by using principal component analysis, which models the\nglobal anatomical relationships between landmarks. We have evaluated PIN using\n72 3D ultrasound images from fetal screening examinations. PIN achieves\nquantitatively an average landmark localisation error of 5.59mm and a runtime\nof 0.44s to predict 10 landmarks per volume. Qualitatively, anatomical 2D\nstandard scan planes derived from the predicted landmark locations are visually\nsimilar to the clinical ground truth. Source code is publicly available at\nhttps://github.com/yuanwei1989/landmark-detection.","url_abs":"http://arxiv.org/abs/1806.06987v2","url_pdf":"http://arxiv.org/pdf/1806.06987v2.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":"fast-multiple-landmark-localisation-using-a","repo_url":"https://github.com/yuanwei1989/landmark-detection","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"multi-task-learning","task_name":"Multi-Task Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}