دورية أكاديمية
أعرض في EDS

Learned iterative segmentation of highly variable anatomy from limited data: Applications to whole heart segmentation for congenital heart disease.

التفاصيل البيبلوغرافية
العنوان: Learned iterative segmentation of highly variable anatomy from limited data: Applications to whole heart segmentation for congenital heart disease.
المؤلفون: Pace DF; Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge, MA, USA; A.A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA. Electronic address: dfpace@mgh.harvard.edu., Dalca AV; Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge, MA, USA; A.A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA., Brosch T; Philips Research Laboratories, Hamburg, Germany., Geva T; Department of Cardiology, Boston Children's Hospital, Boston, MA, USA; Department of Pediatrics, Harvard Medical School, Boston, MA, USA., Powell AJ; Department of Cardiology, Boston Children's Hospital, Boston, MA, USA; Department of Pediatrics, Harvard Medical School, Boston, MA, USA., Weese J; Philips Research Laboratories, Hamburg, Germany., Moghari MH; Department of Cardiology, Boston Children's Hospital, Boston, MA, USA; Department of Pediatrics, Harvard Medical School, Boston, MA, USA., Golland P; Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge, MA, USA.
المصدر: Medical image analysis [Med Image Anal] 2022 Aug; Vol. 80, pp. 102469. Date of Electronic Publication: 2022 May 13.
نوع المنشور: Journal Article; Research Support, Non-U.S. Gov't; Research Support, N.I.H., Extramural
اللغة: English
بيانات الدورية: Publisher: Elsevier Country of Publication: Netherlands NLM ID: 9713490 Publication Model: Print-Electronic Cited Medium: Internet ISSN: 1361-8423 (Electronic) Linking ISSN: 13618415 NLM ISO Abbreviation: Med Image Anal Subsets: MEDLINE
أسماء مطبوعة: Publication: Amsterdam : Elsevier
Original Publication: London : Oxford University Press, [1996-
مواضيع طبية MeSH: Heart Defects, Congenital*/diagnostic imaging , Neural Networks, Computer*, Heart/diagnostic imaging ; Humans ; Image Processing, Computer-Assisted/methods ; Thorax
مستخلص: Training deep learning models that segment an image in one step typically requires a large collection of manually annotated images that captures the anatomical variability in a cohort. This poses challenges when anatomical variability is extreme but training data is limited, as when segmenting cardiac structures in patients with congenital heart disease (CHD). In this paper, we propose an iterative segmentation model and show that it can be accurately learned from a small dataset. Implemented as a recurrent neural network, the model evolves a segmentation over multiple steps, from a single user click until reaching an automatically determined stopping point. We develop a novel loss function that evaluates the entire sequence of output segmentations, and use it to learn model parameters. Segmentations evolve predictably according to growth dynamics encapsulated by training data, which consists of images, partially completed segmentations, and the recommended next step. The user can easily refine the final segmentation by examining those that are earlier or later in the output sequence. Using a dataset of 3D cardiac MR scans from patients with a wide range of CHD types, we show that our iterative model offers better generalization to patients with the most severe heart malformations.
Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
(Copyright © 2022 Elsevier B.V. All rights reserved.)
References: Eur Heart J Cardiovasc Imaging. 2019 Dec 1;20(12):1321-1331. (PMID: 31544926)
IEEE Trans Med Imaging. 2018 Jul;37(7):1562-1573. (PMID: 29969407)
Med Image Anal. 2017 Oct;41:40-54. (PMID: 28526212)
Med Image Comput Comput Assist Interv. 2015 Oct;9351:80-88. (PMID: 26889498)
BMJ. 2016 Jul 28;354:i3905. (PMID: 27470314)
J Thorac Cardiovasc Surg. 2009 Sep;138(3):571-80. (PMID: 19698837)
Med Image Anal. 2017 Jan;35:18-31. (PMID: 27310171)
Med Biol Eng Comput. 2013 Nov;51(11):1209-19. (PMID: 23359255)
Magn Reson Med. 2017 Feb;77(2):759-765. (PMID: 26843458)
Med Image Anal. 2019 Dec;58:101537. (PMID: 31446280)
IEEE Trans Med Imaging. 2008 Nov;27(11):1668-81. (PMID: 18955181)
IEEE Trans Med Imaging. 2010 Sep;29(9):1612-25. (PMID: 20378466)
IEEE Trans Pattern Anal Mach Intell. 2006 Nov;28(11):1768-83. (PMID: 17063682)
Med Image Anal. 2013 Oct;17(7):766-78. (PMID: 23773521)
World J Pediatr Congenit Heart Surg. 2016 May;7(3):344-50. (PMID: 27142402)
Magn Reson Imaging. 2012 Nov;30(9):1323-41. (PMID: 22770690)
IEEE Trans Med Imaging. 2010 Feb;29(2):350-64. (PMID: 19709962)
IEEE Trans Med Imaging. 2008 Sep;27(9):1189-201. (PMID: 18753041)
Proc Natl Acad Sci U S A. 1996 Feb 20;93(4):1591-5. (PMID: 11607632)
Pediatr Cardiol. 2017 Jan;38(1):103-114. (PMID: 27837304)
Med Image Comput Comput Assist Interv. 2011;14(Pt 3):537-45. (PMID: 22003741)
Eur J Cardiothorac Surg. 2017 Dec 1;52(6):1139-1148. (PMID: 28977423)
Neuroimage. 2018 Apr 15;170:434-445. (PMID: 28223187)
Med Image Anal. 2016 Oct;33:38-43. (PMID: 27349830)
Nat Methods. 2018 Aug;15(8):605-610. (PMID: 30013046)
J Am Soc Echocardiogr. 2019 May;32(5):655-666.e13. (PMID: 30826226)
Med Image Anal. 2011 Dec;15(6):863-76. (PMID: 21737337)
J Med Radiat Sci. 2018 Sep;65(3):226-236. (PMID: 29453808)
IEEE Trans Image Process. 2018 Jan 15;:. (PMID: 29994352)
Med Image Anal. 2019 Apr;53:142-155. (PMID: 30771712)
IEEE J Biomed Health Inform. 2019 May;23(3):1151-1162. (PMID: 29994410)
IEEE Trans Med Imaging. 2018 Sep;37(9):2137-2148. (PMID: 29994087)
Neuroimage. 2017 Jul 15;155:159-168. (PMID: 28435096)
IEEE Trans Pattern Anal Mach Intell. 2019 Jul;41(7):1559-1572. (PMID: 29993532)
J Cardiovasc Magn Reson. 2011 Sep 21;13:51. (PMID: 21936913)
Cardiovasc Diagn Ther. 2019 Oct;9(Suppl 2):S310-S325. (PMID: 31737539)
MAGMA. 2016 Apr;29(2):155-95. (PMID: 26811173)
Deep Learn Med Image Anal Multimodal Learn Clin Decis Support (2018). 2018 Sep;11045:334-342. (PMID: 31172133)
Med Image Anal. 2010 Feb;14(1):70-84. (PMID: 19931481)
Pediatr Cardiol. 2018 Mar;39(3):538-547. (PMID: 29181795)
Neuroimage. 2020 Mar;208:116459. (PMID: 31837471)
J Magn Reson Imaging. 2020 Mar;51(3):693-711. (PMID: 31111616)
IEEE Trans Med Imaging. 2011 Sep;30(9):1605-16. (PMID: 21880565)
JRSM Cardiovasc Dis. 2016 Apr 29;5:2048004016645467. (PMID: 27170842)
J Healthc Eng. 2013;4(3):371-408. (PMID: 23965595)
Med Image Anal. 2016 Jul;31:77-87. (PMID: 26999615)
Asian Cardiovasc Thorac Ann. 2018 Jan;26(1):28-37. (PMID: 28185475)
معلومات مُعتمدة: P41 EB015902 United States EB NIBIB NIH HHS; R01 AG064027 United States AG NIA NIH HHS; R01 HL149807 United States HL NHLBI NIH HHS; R56 AG064027 United States AG NIA NIH HHS
فهرسة مساهمة: Keywords: Congenital heart disease; Interactive segmentation; Recurrent Neural Network; Whole heart segmentation
تواريخ الأحداث: Date Created: 20220531 Date Completed: 20220726 Latest Revision: 20240214
رمز التحديث: 20240214
مُعرف محوري في PubMed: PMC9617683
DOI: 10.1016/j.media.2022.102469
PMID: 35640385
قاعدة البيانات: MEDLINE
الوصف
تدمد:1361-8423
DOI:10.1016/j.media.2022.102469