Literature DB >> 31274206

Cross-modality (CT-MRI) prior augmented deep learning for robust lung tumor segmentation from small MR datasets.

Jue Jiang1, Yu-Chi Hu1, Neelam Tyagi1, Pengpeng Zhang1, Andreas Rimner2, Joseph O Deasy1, Harini Veeraraghavan1.   

Abstract

<span class="abstract_title">PURPOSE:pan> Accurate <span class="Disease">tumor segmentation is a requirement for magnetic resonance (MR)-based radiotherapy. Lack of large expert annotated MR datasets makes training deep learning models difficult. Therefore, a cross-modality (MR-CT) deep learning segmentation approach that augments training data using pseudo MR images produced by transforming expert-segmented CT images was developed.
METHODS: Eighty-one T2-weighted MRI scans from 28 patients with non-small cell lung cancers (nine with pretreatment and weekly MRI and the remainder with pre-treatment MRI scans) were analyzed. Cross-modality model encoding the transformation of CT to pseudo MR images resembling T2w MRI was learned as a generative adversarial deep learning network. This model was used to translate 377 expert segmented non-small cell lung cancer CT scans from the Cancer Imaging Archive into pseudo MRI that served as additional training set. This method was benchmarked against shallow learning using random forest, standard data augmentation, and three state-of-the art adversarial learning-based cross-modality data (pseudo MR) augmentation methods. Segmentation accuracy was computed using Dice similarity coefficient (DSC), Hausdorff distance metrics, and volume ratio.
RESULTS: The proposed approach produced the lowest statistical variability in the intensity distribution between pseudo and T2w MR images measured as Kullback-Leibler divergence of 0.069. This method produced the highest segmentation accuracy with a DSC of (0.75 ± 0.12) and the lowest Hausdorff distance of (9.36 mm ± 6.00 mm) on the test dataset using a U-Net structure. This approach produced highly similar estimations of tumor growth as an expert (P = 0.37).
CONCLUSIONS: A novel deep learning MR segmentation was developed that overcomes the limitation of learning robust models from small datasets by leveraging learned cross-modality information using a model that explicitly incorporates knowledge of tumors in modality translation to augment segmentation training. The results show the feasibility of the approach and the corresponding improvement over the state-of-the-art methods.
© 2019 American Association of Physicists in Medicine.

Entities:  

Keywords:  cross-modality learning; data augmentation; generative adversarial networks; magnetic resonance imaging; tumor segmentation

Mesh:

Year:  2019        PMID: 31274206      PMCID: PMC6800584          DOI: 10.1002/mp.13695

Source DB:  PubMed          Journal:  Med Phys        ISSN: 0094-2405            Impact factor:   4.071


  14 in total

1.  On standardizing the MR image intensity scale.

Authors:  L G Nyúl; J K Udupa
Journal:  Magn Reson Med       Date:  1999-12       Impact factor: 4.668

2.  Theil-Sen nonparametric regression technique on univariate calibration, inverse regression and detection limits.

Authors:  Irma Lavagnini; Denis Badocco; Paolo Pastore; Franco Magno
Journal:  Talanta       Date:  2011-10-12       Impact factor: 6.057

3.  Transfer learning improves supervised image segmentation across imaging protocols.

Authors:  Annegreet van Opbroek; M Arfan Ikram; Meike W Vernooij; Marleen de Bruijne
Journal:  IEEE Trans Med Imaging       Date:  2014-11-04       Impact factor: 10.048

Review 4.  The future of image-guided radiotherapy will be MR guided.

Authors:  Julianne M Pollard; Zhifei Wen; Ramaswamy Sadagopan; Jihong Wang; Geoffrey S Ibbott
Journal:  Br J Radiol       Date:  2017-03-29       Impact factor: 3.039

5.  The Future of Artificial Intelligence in Radiation Oncology.

Authors:  Reid F Thompson; Gilmer Valdes; Clifton David Fuller; Colin M Carpenter; Olivier Morin; Sanjay Aneja; William D Lindsay; Hugo J W L Aerts; Barbara Agrimson; Curtiland Deville; Seth A Rosenthal; James B Yu; Charles R Thomas
Journal:  Int J Radiat Oncol Biol Phys       Date:  2018-06-06       Impact factor: 7.038

6.  The Cancer Imaging Archive (TCIA): maintaining and operating a public information repository.

Authors:  Kenneth Clark; Bruce Vendt; Kirk Smith; John Freymann; Justin Kirby; Paul Koppel; Stephen Moore; Stanley Phillips; David Maffitt; Michael Pringle; Lawrence Tarbox; Fred Prior
Journal:  J Digit Imaging       Date:  2013-12       Impact factor: 4.056

7.  Generating synthetic CTs from magnetic resonance images using generative adversarial networks.

Authors:  Hajar Emami; Ming Dong; Siamak P Nejad-Davarani; Carri K Glide-Hurst
Journal:  Med Phys       Date:  2018-06-14       Impact factor: 4.071

8.  New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1).

Authors:  E A Eisenhauer; P Therasse; J Bogaerts; L H Schwartz; D Sargent; R Ford; J Dancey; S Arbuck; S Gwyther; M Mooney; L Rubinstein; L Shankar; L Dodd; R Kaplan; D Lacombe; J Verweij
Journal:  Eur J Cancer       Date:  2009-01       Impact factor: 9.162

9.  Efficient multi-scale 3D CNN with fully connected CRF for accurate brain lesion segmentation.

Authors:  Konstantinos Kamnitsas; Christian Ledig; Virginia F J Newcombe; Joanna P Simpson; Andrew D Kane; David K Menon; Daniel Rueckert; Ben Glocker
Journal:  Med Image Anal       Date:  2016-10-29       Impact factor: 8.545

10.  Tumor delineation: The weakest link in the search for accuracy in radiotherapy.

Authors:  C F Njeh
Journal:  J Med Phys       Date:  2008-10
View more
  7 in total

1.  Random Multi-Channel Image Synthesis for Multiplexed Immunofluorescence Imaging.

Authors:  Shunxing Bao; Yucheng Tang; Ho Hin Lee; Riqiang Gao; Sophie Chiron; Ilwoo Lyu; Lori A Coburn; Keith T Wilson; Joseph T Roland; Bennett A Landman; Yuankai Huo
Journal:  Proc Mach Learn Res       Date:  2021-09

2.  Lung parenchymal characterization via thoracic dynamic MRI in normal children and pediatric patients with TIS.

Authors:  Yubing Tong; Jayaram K Udupa; Joseph M McDonough; Carina Lott; Caiyun Wu; Chamith S Rajapakse; Jason B Anari; Drew A Torigian; Patrick J Cahill
Journal:  Proc SPIE Int Soc Opt Eng       Date:  2021-02-15

Review 3.  A review on medical imaging synthesis using deep learning and its clinical applications.

Authors:  Tonghe Wang; Yang Lei; Yabo Fu; Jacob F Wynne; Walter J Curran; Tian Liu; Xiaofeng Yang
Journal:  J Appl Clin Med Phys       Date:  2020-12-11       Impact factor: 2.102

4.  Computational approaches to detect small lesions in 18 F-FDG PET/CT scans.

Authors:  Kenneth J Nichols; Frank P DiFilippo; Christopher J Palestro
Journal:  J Appl Clin Med Phys       Date:  2021-10-13       Impact factor: 2.102

5.  Teacher-student approach for lung tumor segmentation from mixed-supervised datasets.

Authors:  Vemund Fredriksen; Svein Ole M Sevle; André Pedersen; Thomas Langø; Gabriel Kiss; Frank Lindseth
Journal:  PLoS One       Date:  2022-04-05       Impact factor: 3.240

Review 6.  Deep learning in structural and functional lung image analysis.

Authors:  Joshua R Astley; Jim M Wild; Bilal A Tahir
Journal:  Br J Radiol       Date:  2021-04-20       Impact factor: 3.629

7.  Gross Tumor Volume Segmentation for Stage III NSCLC Radiotherapy Using 3D ResSE-Unet.

Authors:  Xinhao Yu; Fu Jin; HuanLi Luo; Qianqian Lei; Yongzhong Wu
Journal:  Technol Cancer Res Treat       Date:  2022 Jan-Dec
  7 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.