Federated Learning: A Cross-Institutional Feasibility Study of Deep Learning Based Intracranial Tumor Delineation Framework for Stereotactic Radiosurgery.
convolutional neural network
federated learning
magnetic resonance imaging
meningioma
metastasis
vestibular schwannoma
Journal
Journal of magnetic resonance imaging : JMRI
ISSN: 1522-2586
Titre abrégé: J Magn Reson Imaging
Pays: United States
ID NLM: 9105850
Informations de publication
Date de publication:
12 Aug 2023
12 Aug 2023
Historique:
revised:
27
07
2023
received:
18
05
2023
accepted:
28
07
2023
pubmed:
13
8
2023
medline:
13
8
2023
entrez:
12
8
2023
Statut:
aheadofprint
Résumé
Deep learning-based segmentation algorithms usually required large or multi-institute data sets to improve the performance and ability of generalization. However, protecting patient privacy is a key concern in the multi-institutional studies when conventional centralized learning (CL) is used. To explores the feasibility of a proposed lesion delineation for stereotactic radiosurgery (SRS) scheme for federated learning (FL), which can solve decentralization and privacy protection concerns. Retrospective. 506 and 118 vestibular schwannoma patients aged 15-88 and 22-85 from two institutes, respectively; 1069 and 256 meningioma patients aged 12-91 and 23-85, respectively; 574 and 705 brain metastasis patients aged 26-92 and 28-89, respectively. 1.5T, spin-echo, and gradient-echo [Correction added after first online publication on 21 August 2023. Field Strength has been changed to "1.5T" from "5T" in this sentence.]. The proposed lesion delineation method was integrated into an FL framework, and CL models were established as the baseline. The effect of image standardization strategies was also explored. The dice coefficient was used to evaluate the segmentation between the predicted delineation and the ground truth, which was manual delineated by neurosurgeons and a neuroradiologist. The paired t-test was applied to compare the mean for the evaluated dice scores (p < 0.05). FL performed the comparable mean dice coefficient to CL for the testing set of Taipei Veterans General Hospital regardless of standardization and parameter; for the Taichung Veterans General Hospital data, CL significantly (p < 0.05) outperformed FL while using bi-parameter, but comparable results while using single-parameter. For the non-SRS data, FL achieved the comparable applicability to CL with mean dice 0.78 versus 0.78 (without standardization), and outperformed to the baseline models of two institutes. The proposed lesion delineation successfully implemented into an FL framework. The FL models were applicable on SRS data of each participating institute, and the FL exhibited comparable mean dice coefficient to CL on non-SRS dataset. Standardization strategies would be recommended when FL is used. 4 TECHNICAL EFFICACY: Stage 1.
Sections du résumé
BACKGROUND
BACKGROUND
Deep learning-based segmentation algorithms usually required large or multi-institute data sets to improve the performance and ability of generalization. However, protecting patient privacy is a key concern in the multi-institutional studies when conventional centralized learning (CL) is used.
PURPOSE
OBJECTIVE
To explores the feasibility of a proposed lesion delineation for stereotactic radiosurgery (SRS) scheme for federated learning (FL), which can solve decentralization and privacy protection concerns.
STUDY TYPE
METHODS
Retrospective.
SUBJECTS
METHODS
506 and 118 vestibular schwannoma patients aged 15-88 and 22-85 from two institutes, respectively; 1069 and 256 meningioma patients aged 12-91 and 23-85, respectively; 574 and 705 brain metastasis patients aged 26-92 and 28-89, respectively.
FIELD STRENGTH/SEQUENCE
UNASSIGNED
1.5T, spin-echo, and gradient-echo [Correction added after first online publication on 21 August 2023. Field Strength has been changed to "1.5T" from "5T" in this sentence.].
ASSESSMENT
RESULTS
The proposed lesion delineation method was integrated into an FL framework, and CL models were established as the baseline. The effect of image standardization strategies was also explored. The dice coefficient was used to evaluate the segmentation between the predicted delineation and the ground truth, which was manual delineated by neurosurgeons and a neuroradiologist.
STATISTICAL TESTS
METHODS
The paired t-test was applied to compare the mean for the evaluated dice scores (p < 0.05).
RESULTS
RESULTS
FL performed the comparable mean dice coefficient to CL for the testing set of Taipei Veterans General Hospital regardless of standardization and parameter; for the Taichung Veterans General Hospital data, CL significantly (p < 0.05) outperformed FL while using bi-parameter, but comparable results while using single-parameter. For the non-SRS data, FL achieved the comparable applicability to CL with mean dice 0.78 versus 0.78 (without standardization), and outperformed to the baseline models of two institutes.
DATA CONCLUSION
CONCLUSIONS
The proposed lesion delineation successfully implemented into an FL framework. The FL models were applicable on SRS data of each participating institute, and the FL exhibited comparable mean dice coefficient to CL on non-SRS dataset. Standardization strategies would be recommended when FL is used.
LEVEL OF EVIDENCE
METHODS
4 TECHNICAL EFFICACY: Stage 1.
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Subventions
Organisme : Brain Research Center, and the National Yang Ming Chiao Tung University from the Featured Areas Research Center Program within the framework of the Higher Education Sprout Project by the Ministry of Education (MOE) in Taiwan
ID : 112W32101
Organisme : Higher Education Sprout Project of the National Yang Ming Chiao Tung University and Ministry of Education (MOE), Taiwan
Organisme : National Science and Technology Council
ID : 110-2221-E-A49A-504-MY3
Organisme : National Science and Technology Council
ID : 110-2634-F-A49-005
Organisme : National Science and Technology Council
ID : 111-2634-F-006-012
Organisme : National Science and Technology Council
ID : 111-2811-E-A49A-006-MY2
Organisme : National Science and Technology Council
ID : 111-2823-8-A49-001
Organisme : National Science and Technology Council
ID : 112-2823-8-A49-002
Organisme : National Science and Technology Council
ID : 111-2634-F-A49-014
Organisme : Veteran General Hospitals University System of Taiwan
ID : VGHUST110-G7-2-1
Informations de copyright
© 2023 International Society for Magnetic Resonance in Medicine.
Références
Konečný J, McMahan HB, Yu FX, Richtárik P, Suresh AT, Bacon D. Federated learning: Strategies for improving communication efficiency. arXiv:161005492 2016. https://doi.org/10.48550/arXiv.1610.05492
Rieke N, Hancox J, Li W, et al. The future of digital health with federated learning. NPJ Digit Med 2020;3(1):1-7.
Yang Q, Liu Y, Cheng Y, Kang Y, Chen T, Yu H. Federated learning. Synth Lect Artif Intell Mach Learn 2019;13(3):1-207.
Bonawitz K, Eichner H, Grieskamp W, et al. Towards federated learning at scale: System design. Proc Mach Learn Syst 2019;1:374-388.
Li T, Sahu AK, Talwalkar A, Smith V. Federated learning: Challenges, methods, and future directions. IEEE Signal Process Mag 2020;37(3):50-60.
Ng D, Lan X, Yao MM-S, Chan WP, Feng M. Federated learning: A collaborative effort to achieve better medical imaging models for individual sites that have small labelled datasets. Quant Imaging Med Surg 2021;11(2):852-857.
Liu B, Yan B, Zhou Y, Yang Y, Zhang Y. Experiments of federated learning for covid-19 chest x-ray images. arXiv:200705592 2020. https://doi.org/10.48550/arXiv.2007.05592
Roth HR, Chang K, Singh P, et al. Domain adaptation and representation transfer, and distributed and collaborative learning: Second MICCAI Workshop, DART 2020, and First MICCAI Workshop, DCL 2020, Held in Conjunction with MICCAI. Springer; Lima, Peru. 2020; 12444: p 181-191.
Sheller MJ, Edwards B, Reina GA, et al. Federated learning in medicine: Facilitating multi-institutional collaborations without sharing patient data. Sci Rep 2020;10(1):1-12.
NVIDIA Clara. NVIDIA CLARA: An application framework optimized for healthcare and life sciences developers: NVIDIA Clara; 2022. https://developer.nvidia.com/clara
Beutel DJ, Topal T, Mathur A, et al. Flower: A friendly federated learning research framework. arXiv:200714390 2020. https://doi.org/10.48550/arXiv.2007.14390
Ziller A, Trask A, Lopardo A, et al. PySyft: A library for easy federated learning. Studies in Computational Intelligence 2021;111-139.
Sohan MF, Basalamah A. A systematic review on federated learning in medical image analysis. IEEE Access 2023;11:28628-28644.
Luo G, Liu T, Lu J, et al. Influence of data distribution on federated learning performance in tumor segmentation. Radiol Artif Intell 2023;5(3):e220082.
Mehta M, Shao C. Federated learning-based semantic segmentation for pixel-wise defect detection in additive manufacturing. J Manuf Syst 2022;64:197-210.
Naeem A, Anees T, Naqvi RA, Loh W-K. A comprehensive analysis of recent deep and federated-learning-based methodologies for brain tumor diagnosis. J Pers Med 2022;12(2):275.
Yi L, Zhang J, Zhang R, Shi J, Wang G, Liu X. SU-net: An efficient encoder-decoder model of federated learning for brain tumor segmentation. International Conference on Artificial Neural Networks: Springer; 2020. p 761-773.
Sheller MJ, Reina GA, Edwards B, Martin J, Bakas S. Multi-institutional deep learning modeling without sharing patient data: A feasibility study on brain tumor segmentation. International MICCAI Brainlesion Workshop: Springer;Granada, Spain; 2018. p 92-104.
Li W, Milletarì F, Xu D, et al. Privacy-preserving federated brain tumour segmentation. International workshop on machine learning in medical imaging: 10th International Workshop, MLMI 2019, Held in Conjunction with MICCAI. Springer. 2019. p 133-141.
Nyström L. Inter-hospital brain tumour diagnostics using private federated learning an empirical analysis of convergence in a heterogeneous, non-IID setting and a theoretical review of privacy mechanisms. Sweden: Chalmers University of Technology; 2020.
Zhao Y, Li M, Lai L, Suda N, Civin D, Chandra V. Federated learning with non-iid data. arXiv:180600582 2018. https://doi.org/10.48550/arXiv.1806.00582
Liu Q, Dou Q, Yu L, Heng PA. MS-net: Multi-site network for improving prostate segmentation with heterogeneous MRI data. IEEE Trans Med Imaging 2020;39(9):2713-2724.
Lee W-K, Wu C-C, Lee C-C, et al. Combining analysis of multi-parametric MR images into a convolutional neural network: Precise target delineation for vestibular schwannoma treatment planning. Artif Intell Med 2020;107:101911.
Lee C-C, Lee W-K, Wu C-C, et al. Applying artificial intelligence to longitudinal imaging analysis of vestibular schwannoma following radiosurgery. Sci Rep 2021;11(1):3106.
Yang H-C, Wu C-C, Lee C-C, et al. Prediction of pseudoprogression and long-term outcome of vestibular schwannoma after gamma knife radiosurgery based on preradiosurgical MR radiomics. Radiother Oncol 2021;155:123-130.
Lee W-K, Yang H-C, Lee C-C, et al. Lesion delineation framework for vestibular schwannoma, meningioma and brain metastasis for gamma knife radiosurgery using stereotactic magnetic resonance images. Comput Methods Programs Biomed 2023;229:107311.
Speckter H, Santana J, Bido J, et al. Texture analysis of standard magnetic resonance images to predict response to gamma knife radiosurgery in vestibular schwannomas. World Neurosurg 2019;132:e228-e234.
Combs SE, Baumert BG, Bendszus M, et al. ESTRO ACROP guideline for target volume delineation of skull base tumors. Radiother Oncol 2021;156:80-94.
Isensee F, Jaeger PF, Kohl SA, Petersen J, Maier-Hein KH. nnU-net: A self-configuring method for deep learning-based biomedical image segmentation. Nat Methods 2021;18(2):203-211.
Lowekamp BC, Chen DT, Ibáñez L, Blezek D. The design of SimpleITK. Front Neuroinform 2013;7:45.
Ahmad P, Jin H, Qamar S, Zheng R, Jiang W. Combined 3D CNN for brain tumor segmentation. 2020 IEEE conference on multimedia information processing and retrieval (MIPR). 2020 IEEE Conference on Multimedia Information Processing and Retrieval (MIPR) China; 2020. p 109-112.
Anand VK, Grampurohit S, Aurangabadkar P, et al. Brain tumor segmentation and survival prediction using automatic hard mining in 3D CNN architecture. In: Crimi A, Bakas S, editors. Brainlesion: Glioma, multiple sclerosis, stroke and traumatic brain injuries. Cham: Springer International Publishing; 2021. p 310-319.
Rehman A, Khan MA, Saba T, Mehmood Z, Tariq U, Ayesha N. Microscopic brain tumor detection and classification using 3D CNN and feature selection architecture. Microsc Res Tech 2021;84(1):133-149.
Wu Y, He K. Group Normalization. Proc Eur Conf Comput Vis. Proceedings of the European conference on computer vision (ECCV). 2018;3-19.
Sudre CH, Li W, Vercauteren T, Ourselin S, Cardoso MJ. Generalised dice overlap as a deep learning loss function for highly unbalanced segmentations. Deep Learn Med Image Anal Multimodal Learn Clin Decis Support 2017;2017:240-248.
Lin D. An information-theoretic definition of similarity. ICML 1998;98:296-304.
Wang B, Lei Y, Tian S, et al. Deeply supervised 3D fully convolutional networks with group dilated convolution for automatic MRI prostate segmentation. Med Phys 2019;46(4):1707-1718.
Isensee F, Schell M, Pflueger I, et al. Automated brain extraction of multisequence MRI using artificial neural networks. Hum Brain Mapp 2019;40(17):4952-4964.