Federated learning improves site performance in multicenter deep learning without data sharing
Open accessAbstract
Objective: To demonstrate enabling multi-institutional training without centralizing or sharing the underlying physical data via federated learning (FL). Materials and Methods: Deep learning models were trained at each participating institution using local clinical data, and an additional model was trained using FL across all of the institutions. Results: We found that the FL model exhibited superior performance and generalizability to the models trained at single institutions, with an overall performance level that was significantly better than that of any of the institutional models alone when evaluated on held-out test sets from each institution and an outside challenge dataset. Discussion: The power of FL was successfully demonstrated across 3 academic institutions while avoiding the privacy risk associated with the transfer and pooling of patient data. Conclusion: Federated learning is an effective methodology that merits further study to enable accelerated development of models across institutions, enabling greater generalizability in clinical use.
AI snapshot
Written from the paper’s full text.
TL;DR
In a 3-institution study using real prostate MRI, federated learning produced a single model without sharing patient data. The FL model outperformed each institution’s private model on held-out institutional test sets and an external challenge dataset.
Key findings
- The FL model achieved an overall private-test-set mean Dice of 0.895 ± 0.036, significantly better than each private model (P < .001); private model overall Dice values were 0.833 ± 0.131 (NCI), 0.854 ± 0.074 (SUNY), and 0.745 ± 0.178 (UCLA).
- On the external PROSTATEx challenge dataset (n = 343), the FL model achieved 0.889 ± 0.036, significantly better than the private models: NCI 0.872 ± 0.062, SUNY 0.838 ± 0.043, UCLA 0.812 ± 0.136 (P < .001 for all comparisons).
- Private models performed well on their own institutional test sets (Dice range 0.883–0.925) but had diminished performance on other institutions’ test sets (Dice range 0.575–0.887).
Methods
- Retrospective multicenter study at UCLA, SUNY Upstate, and NCI: each institution collected 1 prostate MRI from 100 patients, with axial T2W images at 3T and expert whole-prostate segmentations; each private dataset was split into 80 training and 20 held-out test images. An external challenge dataset of 343 axial T2W prostate images from SPIE-AAPM-NCI PROSTATEx was also used.
- The model was a 3D Anisotropic Hybrid Network trained with soft Dice loss and Adam; each private training set was split into 5 sets of 16 images to train 5 submodels combined into an ensemble. All models were trained for 300 epochs.
- Federated learning used a cloud-based aggregation server; after each epoch, model weights and validation metrics were sent from each institution, combined by a weighted average based on validation mean Dice, and redistributed. Evaluation used Dice coefficient and 2-sided paired t-tests comparing private models to the FL model on the combined private test set and the challenge dataset.
Limitations
- Stated: the authors did not address the potential for an inside actor to recover underlying patient data through a model inversion attack on shared weights; they suggest calibrated distortion to shared weights as a possible future enhancement.
- Stated: the task (prostate segmentation on T2-weighted MRI) is relatively simple, and all private models achieved high performance on their own institutional datasets, so the study could not demonstrate that an FL-trained model would significantly outperform a single-site-trained model on that single site’s data.
- Stated: because private datasets were similarly sized, the authors did not explore varying the federated aggregation methodology; adding more institutions may introduce challenges in imaging data quality heterogeneity, governance, intellectual property, and model generalizability, possibly requiring private fine-tuning. Evident from the text: only 3 academic institutions and one segmentation task were tested, and no formal privacy attack evaluation was performed.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
4 of 9 cited sources checked · 2 supported · 0 flagged · accuracy not yet measured
- Premisecitation
Works that examine the practical application of FL in radiological applications are still limited.
“However, works that examine the practical application of FL in radiological applications are still limited.20,21”
Quote in the paperSubstantial claim1 of 2 sources checked- cites [20] Roth
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- cites [21] Sheller, 2020, Federated learning in medicine: facilitating multi-institutional collaborations without sharing patient data, Sci Rep source
The Methods > Data and Discussion passages describe this paper’s evaluation of federated learning on radiographic brain imaging, so it is an example of the kind of work the citing sentence identifies. They do not establish that such works are limited in number. No page numbers are available in our copy. (LLM, conf=0.81)
Experimental — accuracy not yet measured
- Findingown data
The private models exhibited varied performance on the challenge dataset (Dice 0.812-0.872).
“The private models exhibited varied performance on the challenge dataset (Dice coefficient range: 0.812–0.872).”
Quote in the paperNot a substantial claim - Findingreasoning
The FL model did not significantly outperform a single-site-trained model on that single site's data.
“we were unable to demonstrate the expected benefit that an FL-trained model would significantly outperform a single-site-trained model on that single site’s data.”
Quote in the paperSubstantial claim - Study design and conductreasoning
No private data were transferred or shared across institutions during the study.
“At no point during this study were private data transferred or shared across institutions.”
- Also stated in: § DISCUSSION · plain
“Since no transfer of protected health information (or even deidentified health information) was required, we were able to address the privacy and data governance limitations inherent to multicenter studies through the use of simplified 2-way collaboration agreements, rather than requiring the negotiation of a complex 4-way collaboration and material transfer agreement that would have been required”
Quote in the paperSubstantial claim - Also stated in: § DISCUSSION · plain
- Interpretationown data
The power of federated learning was successfully demonstrated across 3 academic institutions using real clinical prostate imaging data.
“The power of federated learning was successfully demonstrated across 3 academic institutions using real clinical prostate imaging data.”
Quote in the paperSubstantial claim
- Premisecitation
Deep learning models require a large volume of training data that captures the full breadth of inputs on which the model is likely to be used.
“A major limitation of the DL approach is the need for a large volume of training data that captures the full breadth of inputs on which the model is likely to be subsequently used.”
Quote in the paperSubstantial claim0 of 1 sources checked- cites [13] Russakovsky, 2015, ImageNet large scale visual recognition challenge, Int J Comput Vis source
- Interpretationreasoning
The method demonstrated in the paper significantly better protects patient privacy than the current standard of direct sharing of data.
“we believe the method we demonstrate in this paper significantly better protects the privacy of patients than the current standard of direct sharing of data.”
Quote in the paperSubstantial claim- Rests on an assumption: The practical risk of model inversion by an inside actor outside crafted malintent is low due to the weight averaging scheme.
- Interpretationreasoning
The FL model was able to benefit from learning institution-specific knowledge through the FL aggregation paradigm without any individual training site seeing the full breadth of inputs.
“This suggests that the FL model was able to benefit from the advantage of learning important institution-specific knowledge through the FL aggregation paradigm, without requiring any individual training site to “see” the full breadth of inputs.”
Quote in the paperSubstantial claim - Study design and conductnone given
Participating institutions intentionally made no effort to harmonize either the T2W acquisition protocol or segmentation methodologies.
“participating institutions intentionally made no effort to harmonize either the T2W acquisition protocol or segmentation methodologies.”
Quote in the paperSubstantial claim - Interpretationown data
The FL model attained the expected advantages inherent in training with more data through the FL aggregation method, even though the full dataset was not seen at any single training site.
“suggesting that the model also attained the expected advantages inherent in training with more data through the FL aggregation method, even though the full dataset was not seen at any single training site.”
Quote in the paperSubstantial claim
- Classified as probably not a substantial claim. 6
- May combine two claims. 8
Probably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationPrevious work has demonstrated the potential utility of FL for model training, generally using publicly available data to simulate multi-institutional training.
“Previous work has demonstrated the potential utility of FL for model training, generally using publicly available data to simulate multi-institutional training.”
Quote in the paperNot a substantial claim1 of 2 sources checked- cites [20] Roth
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- cites [21] Sheller, 2020, Federated learning in medicine: facilitating multi-institutional collaborations without sharing patient data, Sci Rep source
May combine two claims.
PremisecitationDeep learning models in medical imaging are often trained on single-institution datasets and often suffer from poor performance when transferred to other institutions with differing protocols, equipment, or patient populations.
“Such models often suffer from poor performance when transferred to other institutions with differing protocols, equipment, or patient populations.14,15”
Quote in the paperSubstantial claim2 of 2 sources checked- cites [14] Yasaka, 2018, Deep learning and artificial intelligence in radiology: current applications and future directions, PLOS Med source
- cites [15] De Fauw, 2018, Clinically applicable deep learning for diagnosis and referral in retinal disease, Nat Med source
The Introduction (p. 1) identifies generalization to new populations and devices as a challenge. In “Generalization to a New Scanning Device Type” (p. 7), the authors report a 46.6% referral-error rate when their model was applied to scans from a different device without retraining. That supports the equipment-transfer point, but the study does not establish that models often perform poorly across institutions or that differing protocols or patient populations caused such declines. (LLM, conf=0.88)
Experimental — accuracy not yet measured
May combine two claims.
Findingown dataThe private models performed well on their own private test sets (Dice 0.883-0.925) but had diminished performance on other private test sets (Dice 0.575-0.887), with overall mean Dice coefficients between 0.745 and 0.833.
“The private models performed well on their own private test sets (Dice coefficient range: 0.883–0.925) but had diminished performance on the other private test sets (Dice coefficient range: 0.575–0.887). This led to overall mean Dice coefficients between 0.745 and 0.833 for the private models.”
Quote in the paperSubstantial claimMay combine two claims.
Findingown dataOn the external ProstateX challenge dataset (n=343), the FL model achieved an overall mean Dice coefficient of 0.889 and was significantly superior to each private model (P < .001).
“The generic FL model outperformed each of the private models, with an overall mean Dice coefficient of 0.889. The statistical analysis again demonstrated that the FL model was significantly superior to any of the private models (P < .001).”
Quote in the paperSubstantial claimMay combine two claims.
Findingown dataOn the combined private test sets (n=60), the FL model achieved an overall mean Dice coefficient of 0.895 (range 0.880-0.920) and was significantly superior to each private model (P < .001 for all comparisons).
“The FL model exhibited private test set mean Dice coefficients between 0.880 and 0.920, yielding an overall result of 0.895. The statistical analysis using 2-sided paired t-tests demonstrated that the FL model was significantly superior to any of the private models (P < .001 for all comparisons).”
- Also stated in: § DISCUSSION · plain
“The FL model that we trained performed well across all of the private datasets, yielding an overall performance level that was significantly better than that of any of the private models alone.”
Quote in the paperSubstantial claim- Also stated in: § DISCUSSION · plain
May combine two claims.
Interpretationown dataThe federated model demonstrated improved performance across both held-out test sets from each institution and an external test set, validating the FL paradigm.
“The federated model demonstrated improved performance across both held-out test sets from each institution and an external test set, validating the FL paradigm.”
Quote in the paperSubstantial claimMay combine two claims.
InterpretationreasoningIn the worst case, the method is no less secure than current practice because only model weights are ever transmitted.
“in the worst case, our method is no less secure as only model weights are ever transmitted.”
Quote in the paperSubstantial claimProbably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.
Study design and conductnone givenAll 3 academic institutions had IRB approval for review and image analysis, with written informed patient consent or waiver of patient consent.
“All 3 academic institutions had IRB approval for review and image analysis, with written informed patient consent or waiver of patient consent.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
Study design and conductown dataThe study did not attempt to address the potential for an inside actor to recover underlying patient data through a model inversion attack on the trained weights shared during federated learning.
“we did not attempt to address the potential for an inside actor (ie, 1 of the participating institutions) to attempt to recover the underlying patient data through a model inversion attack on the trained weights shared during federated learning.”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
PremisecitationThe sharing of trained model weights is an accepted practice within healthcare.
“the sharing of trained model weights is an accepted practice within healthcare,28,29”
Quote in the paperNot a substantial claim0 of 2 sources checked- cites [28] Soni
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
- cites [29] Liu, 2020
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
May combine two claims.
Interpretationown dataFederated learning can be reduced to practice using real-world private clinical data across multiple institutions, creating a model with improved generalizability both within participating institutions and with outside data.
“Our work shows that FL can be reduced to practice using real-world private clinical data across multiple institutions, and that this approach creates a model that demonstrates improved generalizability both within the participating institutions and with outside data.”
- Also stated in: § Results · plain
“We found that the FL model exhibited superior performance and generalizability to the models trained at single institutions, with an overall performance level that was significantly better than that of any of the institutional models alone when evaluated on held-out test sets from each institution and an outside challenge dataset.”
- Also stated in: § CONCLUSION · boosted
“The federated model demonstrated improved performance across both held-out test sets from each institution and an external test set, validating the FL paradigm.”
Quote in the paperSubstantial claim- Also stated in: § Results · plain
Probably not a substantial claim: may not pass S3 (load-bearing).
Interpretationnone givenModel inversion is a technical risk that cannot be ruled out.
“though model inversion is a technical risk that cannot be ruled out”
Quote in the paperNot a substantial claimProbably not a substantial claim: may not pass S3 (load-bearing).
PremisereasoningA limitation of FL is that no single model ever sees a complete picture of the universe of potential inputs during training, placing pressure on the federated aggregation function to adequately distribute knowledge from each site into the model.
“However, a limitation of this approach is that no single model ever “sees” a complete picture of the universe of potential inputs during the training phase, thus placing pressure on the federated aggregation function to adequately distribute knowledge from each site into the model.”
Quote in the paperNot a substantial claim
The paper relies on this without support.
The practical risk of model inversion by an inside actor outside crafted malintent is low due to the weight averaging scheme.
“we empirically believe that the practical risk of inversion outside of crafted malintent on the part of study designers to be low due to the weight averaging scheme in place.”
The paper relies on this without support.
Medical imaging data in most cases is siloed within provider institutions, and assembling large-scale datasets traditionally requires transfer of data between silos.
“However, medical imaging data in most cases is siloed within provider institutions, and, as a result, assembling large-scale datasets traditionally requires the transfer of data between these silos.”
The paper relies on this without support.
Very few public large-scale pooled medical image datasets exist.
“As a result, very few public large-scale pooled medical image datasets exist.”
The paper relies on this without support.
Deep learning models are primarily interpolators, not extrapolators—they perform best when presented with inputs similar to training data.
“This large volume is required because deep learning models are primarily interpolators, not extrapolators—that is, they perform best when presented with inputs that are similar to the data that those models were trained on.”
The paper relies on this without support.
The task of prostate segmentation on T2-weighted MRI is relatively simple.
“the task we used (prostate segmentation on T2-weighted MRI) is relatively simple”
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Federated learning provides an alternative paradigm for accurate and generalizable machine learning by only sharing numerical model updates from decentralized data.
“Federated ML (FL) provides an alternative paradigm for accurate and generalizable ML, by only sharing numerical model updates.”
From Federated learning enables big data for rare cancer boundary detection (2022) · cites this paper as [19]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Private-model overall range omits the SUNY result
MinorThe stated upper bound is 0.833, but Table 3 reports an overall mean Dice coefficient of 0.854 for the SUNY private model.
“This led to overall mean Dice coefficients between 0.745 and 0.833 for the private models.”
“| Private models | SUNY | 0.887 ± 0.027 | 0.906 ± 0.018 | 0.768 ± 0.064 | 0.854 ± 0.074* |”
Cites
ImageNet Large Scale Visual Recognition Challenge
Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., Huang, Z., Karpathy, A., Khosla, A., Bernstein, M., et al. (2015) · International Journal of Computer Vision · doi:10.1007/s11263-015-0816-y · Cited by 39,380 · cited 1× · see citing works
“13 This large volume is required because deep learning models are primarily interpolators , not extrapolators— that is, they perform best when presented with inputs that are similar to the data that those models were trained on.”
Dermatologist-level classification of skin cancer with deep neural networks
Esteva, A., Kuprel, B., Novoa, R. A., Ko, J., Swetter, S. M., Blau, H. M., & Thrun, S. (2017) · Nature · doi:10.1038/nature21056 · Cited by 14,362 · see citing works
Gulshan, V., Peng, L., Coram, M., Stumpe, M. C., Wu, D., Narayanaswamy, A., Venugopalan, S., Widner, K., Madams, T., Cuadros, J., et al. (2016) · JAMA · doi:10.1001/jama.2016.17216 · Cited by 7,697 · cited 1× · see citing works
“DL has enabled a wave of innovation in imaging decision support, with recent major results in the fields of ophthalmology, 2–4 dermatology, 5–7 pathology, 8–10 and radiology.”
Li, Z., & Hoiem, D. (2018) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2017.2773081 · Cited by 4,202 · see citing works
The Future of Digital Health with Federated Learning
Rieke, N., Hancox, J., Li, W., Milletari, F., Roth, H., Albarqouni, S., Bakas, S., Galtier, M. N., Landman, B., Maier-Hein, K., et al. (2020) · arXiv · doi:10.48550/arxiv.2003.08119 · Cited by 3,495
Self-citationEhteshami Bejnordi, B., Veta, M., Johannes van Diest, P., van Ginneken, B., Karssemeijer, N., Litjens, G., van der Laak, J. A. W. M., and the CAMELYON16 Consortium, Hermsen, M., Manson, Q. F., et al. (2017) · JAMA · doi:10.1001/jama.2017.14585 · Cited by 3,293 · see citing works
International evaluation of an AI system for breast cancer screening
McKinney, S. M., Sieniek, M., Godbole, V., Godwin, J., Antropova, N., Ashrafian, H., Back, T., Chesus, M., Corrado, G. S., Darzi, A., et al. (2020) · Nature · doi:10.1038/s41586-019-1799-6 · Cited by 3,287 · cited 1× · see citing works
“DL has enabled a wave of innovation in imaging decision support, with recent major results in the fields of ophthalmology, dermatology, pathology, and radiology.(11,12) A major limitation of the DL approach is the need for a large volume of training data that captures the full breadth of inputs on which the model is l…”
BackgroundCoudray, N., Ocampo, P. S., Sakellaropoulos, T., Narula, N., Snuderl, M., Fenyö, D., Moreira, A. L., Razavian, N., & Tsirigos, A. (2018) · Nature Medicine · doi:10.1038/s41591-018-0177-5 · Cited by 2,930 · see citing works
Clinically applicable deep learning for diagnosis and referral in retinal disease
De Fauw, J., Ledsam, J. R., Romera-Paredes, B., Nikolov, S., Tomasev, N., Blackwell, S., Askham, H., Glorot, X., O’Donoghue, B., Visentin, D., et al. (2018) · Nature Medicine · doi:10.1038/s41591-018-0107-6 · Cited by 2,639 · cited 1× · see citing works
“14,15 As a result, there is a need for methods to enable the development of general models for clinical use, without requiring the creation of pooled datasets.”
BackgroundSheller, M. J., Edwards, B., Reina, G. A., Martin, J., Pati, S., Kotrotsou, A., Milchenko, M., Xu, W., Marcus, D., Colen, R. R., et al. (2020) · Scientific Reports · doi:10.1038/s41598-020-69250-1 · Cited by 1,646 · cited 2× · see citing works
“21 In FL, models are trained simultaneously at each site and then periodically aggregated and redistrib-uted.”
BackgroundMethodSecure, privacy-preserving and federated machine learning in medical imaging
Kaissis, G. A., Makowski, M. R., Rückert, D., & Braren, R. F. (2020) · Nature Machine Intelligence · doi:10.1038/s42256-020-0186-1 · Cited by 1,501 · see citing works
Deep learning algorithms for detection of critical findings in head CT scans: a retrospective study
Chilamkurthy, S., Ghosh, R., Tanamala, S., Biviji, M., Campeau, N. G., Venugopal, V. K., Mahajan, V., Rao, P., & Warier, P. (2018) · The Lancet · doi:10.1016/s0140-6736(18)31645-3 · Cited by 1,020 · cited 1× · see citing works
“11,12 A major limitation of the DL approach is the need for a large volume of training data that captures the full breadth of inputs on which the model is likely to be subsequently used.”
BackgroundBulten, W., Pinckaers, H., van Boven, H., Vink, R., de Bel, T., van Ginneken, B., van der Laak, J., Hulsbergen-van de Kaa, C., & Litjens, G. (2020) · The Lancet Oncology · doi:10.1016/s1470-2045(19)30739-9 · Cited by 705 · cited 1× · see citing works
“DL has enabled a wave of innovation in imaging decision support, with recent major results in the fields of ophthalmology, 2–4 dermatology, 5–7 pathology, 8–10 and radiology.”
Automatic skin lesion segmentation using deep fully convolutional networks with jaccard distance.
Yuan, Y., Chao, M., & Lo, Y.-C. (2017) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2017.2695227 · Cited by 687 · see citing works
Privacy-Preserving Federated Brain Tumour Segmentation
Li, W., Milletarì, F., Xu, D., Rieke, N., Hancox, J., Zhu, W., Baust, M., Cheng, Y., Ourselin, S., Cardoso, M. J., et al. (2019) · Lecture Notes in Computer Science · doi:10.1007/978-3-030-32692-0_16 · Cited by 578 · see citing works
Self-citationImageNet Large Scale Visual Recognition Challenge
Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., Huang, Z., Karpathy, A., Khosla, A., Bernstein, M., et al. (2014) · arXiv · doi:10.48550/arxiv.1409.0575 · Cited by 567 · see citing works
Zhang, L., Wang, X., Yang, D., Sanford, T., Harmon, S. A., Türkbey, B., Wood, B. J., Roth, H. R., Myronenko, A., Xu, D., et al. (2020) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2020.2973595 · Cited by 556 · see citing works
Self-citationDeep image mining for diabetic retinopathy screening
Quellec, G., Charrière, K., Boudi, Y., Cochener, B., & Lamard, M. (2017) · Medical Image Analysis · doi:10.1016/j.media.2017.04.012 · Cited by 526 · see citing works
Skin lesion classification with ensembles of deep convolutional neural networks
Harangi, B. (2018) · Journal of Biomedical Informatics · doi:10.1016/j.jbi.2018.08.006 · Cited by 451 · see citing works
Distributed deep learning networks among institutions for medical imaging
Chang, K., Balachandar, N., Lam, C., Yi, D., Brown, J., Beers, A., Rosen, B., Rubin, D. L., & Kalpathy-Cramer, J. (2018) · Journal of the American Medical Informatics Association · doi:10.1093/jamia/ocy017 · Cited by 340 · cited 1× · see citing works
“16 In this paradigm, data are not combined into a single, pooled dataset.”
Larobina, M., & Murino, L. (2013) · Journal of Imaging Informatics in Medicine · doi:10.1007/s10278-013-9657-9 · Cited by 244 · cited 1× · see citing works
“The images were then converted to the NIFTI format 23 for training, and the intensity values within each image were normalized to zero mean and unit variance.”
MethodDeep learning and artificial intelligence in radiology: current applications and future directions
Yasaka, K., & Abe, O. (2018) · PLoS Medicine · doi:10.1371/journal.pmed.1002707 · Cited by 238 · cited 1× · see citing works
“Such models often suffer from poor performance when transferred to other institutions with differing protocols, equipment, or patient populations.(14,15) As a result, there is a need for methods to enable the development of general models for clinical use, without requiring the creation of pooled datasets.”
BackgroundCode of Medical Ethics of the American Medical Association
Riddick, F. A. (2011) · Medicine & Science in Sports & Exercise · doi:10.1249/mss.0b013e318215e9ba · Cited by 229 · see citing works
Armato, S. G., Huisman, H. J., Drukker, K., Hadjiiski, L. M., Kirby, J., Petrick, N., Redmond, G., Giger, M. L., Cha, K., Mamonov, A. B., et al. (2018) · Journal of medical imaging · doi:10.1117/1.jmi.5.4.044501 · Cited by 196 · cited 1× · see citing works
“In addition, 343 axial T2W images of the prostate were obtained from the public SPIE-AAPM-NCI PROSTATEx data-set 22 (the “challenge dataset”).”
MethodLiu, S., Xu, D., Zhou, S. K., Pauly, O., Grbić, S., Mertelmeier, T., Wicklein, J., Jerebko, A. K., Cai, W., & Comaniciu, D. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-00934-2_94 · Cited by 129 · cited 1× · see citing works
“The 3D Anisotropic Hybrid Network 24”
Self-citationBulten, W., Pinckaers, H., van Boven, H., Vink, R., de Bel, T., van Ginneken, B., van der Laak, J. A. W. M., van de Kaa, C. A. H. ‐., & Litjens, G. J. S. (2020) · arXiv (Cornell University) · Cited by 106 · see citing works
Accounting for data variability in multi-institutional distributed deep learning for medical imaging
Balachandar, N., Chang, K., Kalpathy–Cramer, J., & Rubin, D. L. (2020) · Journal of the American Medical Informatics Association · doi:10.1093/jamia/ocaa017 · Cited by 61 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
Algorithmic fairness in artificial intelligence for medicine and healthcare
Chen, R. J., Wang, J. J., Williamson, D. F. K., Chen, T. Y., Lipkova, J., Lu, M. Y., Sahai, S., & Mahmood, F. (2023) · Nature Biomedical Engineering · doi:10.1038/s41551-023-01056-8 · Cited by 750 · see citing works
“…multi-institutional collaboration to aid the validation of AI algorithms for prostate segmentation, brain-cancer detection, the monitoring of the progression of Alzheimer’s disease using magnetic resonance imaging (MRI) and the classification of paediatric chest X-ray images 211,214,222–224 .”
BackgroundShifting machine learning for healthcare from development to deployment and from models to data
Zhang, A., Xing, L., Zou, J., & Wu, J. C. (2022) · Nature Biomedical Engineering · doi:10.1038/s41551-022-00898-y · Cited by 420 · see citing works
“…learning were found to achieve 99% of the performance of traditional centralized training even with imbalanced datasets or with relatively few samples per institution, thus showing that federated learning can be realistically implemented without sacrificing performance or generalization 77,78 .”
BackgroundFederated learning enables big data for rare cancer boundary detection
Pati, S., Baid, U., Edwards, B., Sheller, M., Wang, S.-H., Reina, G. A., Foley, P., Gruzdev, A., Karkada, D., Davatzikos, C., et al. (2022) · Nature Communications · doi:10.1038/s41467-022-33407-5 · Cited by 394 · see citing works
“The exact 49 site IDs are: 1, 2, 3, 4, 5, 6, 7, 9, 10, 12, 13, 14, 15, 16, 17, 18, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 44, 45, 46, 48, 49, 50, 52, 53, 54, 56, 59, 60.”
Rauniyar, A., Hagos, D. H., Jha, D., Håkegård, J. E., Bagci, U., Rawat, D. B., & Vlassov, V. (2024) · IEEE Internet of Things Journal · doi:10.1109/jiot.2023.3329061 · Cited by 315 · see citing works
“In [153], the efficacy of FL has been exemplified through a multi-institutional collaboration involving three esteemed academic establishments: 1) the University of California at Los Angeles (UCLA), USA; 2) the State University of New York (SUNY) Upstate Medical University, USA; and 3) the National…”
BackgroundFederated learning for secure IoMT-applications in smart healthcare systems: A comprehensive review
Rani, S., Kataria, A., Kumar, S., & Tiwari, P. (2023) · Knowledge-Based Systems · doi:10.1016/j.knosys.2023.110658 · Cited by 228 · see citing works
Prayitno, Shyu, C.-R., Putra, K. T., Chen, H.-C., Tsai, Y.-Y., Hossain, K. S. M. T., Jiang, W., & Shae, Z.-Y. (2021) · Applied Sciences · doi:10.3390/app112311191 · Cited by 210 · see citing works
“There are several published FL studies in medical image segmentation, namely brain tumor disease [38], COVID-19 region [53], and prostate cancer region [50].”
InfluentialFederated Learning for Healthcare Domain - Pipeline, Applications and Challenges
Joshi, M., Pal, A., & Sankarasubbu, M. (2022) · ACM Transactions on Computing for Healthcare · doi:10.1145/3533708 · Cited by 218 · see citing works
“Each hospital distributes the local model that has been trained with patient data and acquires the global model for future diagnosis [97].”
BackgroundMethodInfluentialApplications of Federated Learning; Taxonomy, Challenges, and Research Trends
Shaheen, M., Farooq, M. S., Umer, T., & Kim, B.-S. (2022) · Electronics · doi:10.3390/electronics11040670 · Cited by 157 · see citing works
Linardos, A., Kushibar, K., Walsh, S., Gkontra, P., & Lekadir, K. (2022) · Scientific Reports · doi:10.1038/s41598-022-07186-4 · Cited by 130 · see citing works
“Privacy preserving federated learning systems for medical image analysis have been mainly explored in the context of segmentation for brain 5–8 , prostate 9 and COVID-19 affected regions 10–12 .”
BackgroundDo Gradient Inversion Attacks Make Federated Learning Unsafe?
Hatamizadeh, A., Yin, H., Molchanov, P. A., Myronenko, A., Li, W., Dogra, P., Feng, A., Flores, M. G., Kautz, J., Xu, D., et al. (2023) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2023.3239391 · Cited by 133 · see citing works
“in scanner equipment and acquisition protocols, often spanning different countries, continents, and patient populations [3], [4], [5], [6], [7], [8].”
BackgroundSelf-citationIslam, M., Reza, M. T., Kaosar, M. G., & Parvez, M. Z. (2022) · Neural Processing Letters · doi:10.1007/s11063-022-11014-1 · Cited by 120 · see citing works
“Federated Learning (FL) is a recently proposed machine learning technique, which tries to address the problem of strict regulations regarding data secrecy and limited supply of available data sets [11, 33, 34].”
BackgroundReviewing Federated Machine Learning and Its Use in Diseases Prediction
Moshawrab, M., Adda, M., Bouzouane, A., Ibrahim, H., & Raad, A. (2023) · Sensors · doi:10.3390/s23042112 · Cited by 114 · see citing works
“Cancer Yes No Yes No [132] Same No Yes No [134] Same No Yes No [135] Same No Yes No [136] Same No Yes No [138] Yes No Yes No [139] Yes No Yes No [140] Not available No Yes No [141] Yes No Yes No [142] Not available No Yes No [143] No No Yes No [144] No No Yes No [145] Not available No Yes No”
BackgroundResultInfluentialGenomics and multiomics in the age of precision medicine
Mani, S., Lalani, S., & Pammi, M. (2025) · Pediatric Research · doi:10.1038/s41390-025-04021-0 · Cited by 114 · see citing works
“81 Application programming interface (API) and related Substitutable Medical Apps, Reusable Technologies (SMART ® ) platform are showing promise in this process of integration.”
Federated learning: Applications, challenges and future directions
Bharati, S., Mondal, M. R. H., Podder, P., & Prasath, V. B. S. (2022) · International Journal of Hybrid Intelligent Systems · doi:10.3233/his-220006 · Cited by 117 · see citing works
“For example, the classification tasks of COVID-19 detections [71–84], cancer diagnoses [85–94] and autism spectrum disorder (ASD) [84,95–97] are considered in a FL setting in healthcare.”
Explainable, Domain-Adaptive, and Federated Artificial Intelligence in Medicine
Chaddad, A., Lu, Q., Li, J., Katib, Y., Kateb, R., Tanougast, C., Bouridane, A., & Abdulkadir, A. (2023) · IEEE/CAA Journal of Automatica Sinica · doi:10.1109/jas.2023.123123 · Cited by 112 · see citing works
“This concept of data-private collaborative learning [51], that Source data”
BackgroundEdge AI and Blockchain for Smart Sustainable Cities: Promise and Potential
Badidi, E. (2022) · Sustainability · doi:10.3390/su14137609 · Cited by 94 · see citing works
The importance of resource awareness in artificial intelligence for healthcare
Jia, Z., Chen, J., Xu, X., Kheir, J. N., Hu, J., Xiao, H., Peng, S., Hu, X. S., Chen, D., & Shi, Y. (2023) · Nature Machine Intelligence · doi:10.1038/s42256-023-00670-0 · Cited by 85 · see citing works
Application of artificial intelligence in cataract management: current and future directions
Gutiérrez, L., Lim, J. S., Foo, L. L., Ng, W. Y., Yip, M. Y. T., Lim, G. Y. S., Wong, M. M., Fong, A., Rosman, M., Mehta, J. S., et al. (2022) · Eye and Vision · doi:10.1186/s40662-021-00273-z · Cited by 83 · see citing works
“This is a privacy-preserving technique that exposes the model to heterogenous non-independent and identically distributed data [66].”
MethodMedical Imaging Applications of Federated Learning
Sandhu, S. S., Gorji, H. T., Tavakolian, P., Tavakolian, K., & Akhbardeh, A. (2023) · Diagnostics · doi:10.3390/diagnostics13193140 · Cited by 90 · see citing works
“Additionally, Sarma et al. conducted a case study demonstrating their ability to utilize FL to train a deep learning model across three academic institutions while preserving patients’ privacy [92].”
Nguyen, T., Dakka, M. A., Diakiw, S. M., VerMilyea, M. D., Perugini, M., Hall, J. M. M., & Perugini, D. (2022) · Scientific Reports · doi:10.1038/s41598-022-12833-x · Cited by 78 · see citing works
“The field of federated learning has rapidly expanded into the area of healthcare 17–21 , in medical applications in particular 22–25 bringing a wide range of methods for AI training across distributed devices or data silos (horizontal or sample-based federated learning), data centers within an…”
BackgroundSunoqrot, M. R. S., Saha, A., Hosseinzadeh, M., Elschot, M., & Huisman, H. (2022) · European Radiology Experimental · doi:10.1186/s41747-022-00288-8 · Cited by 76 · see citing works
“A successful example of federated training of a prostate segmentation algorithm is reported by Sarma et al. [58].”
MethodMachine learning in prostate MRI for prostate cancer: current status and future opportunities
Li, H., Lee, C. H., Chia, D., Lin, Z., Huang, W., & Tan, C. H. (2022) · Diagnostics · doi:10.3390/diagnostics12020289 · Cited by 74 · see citing works
Federated disentangled representation learning for unsupervised brain anomaly detection
Bercea, C. I., Wiestler, B., Rueckert, D., & Albarqouni, S. (2022) · Nature Machine Intelligence · doi:10.1038/s42256-022-00515-2 · Cited by 99 · see citing works
“Second, we augment the original images with up to 3 rectangles of random size s ∈ [ 6 , 500 ] with the width w ∈ [ 3 , 20 ] and height h ∈ [ w − w / 3 , w + w / 3 ] .”
MethodOvercoming data scarcity in biomedical imaging with a foundational multi-task model
Schäfer, R., Nicke, T., Höfener, H., Lange, A., Merhof, D., Feuerhake, F., Schulz, V., Lotz, J., & Kiessling, F. M. (2024) · Nature Computational Science · doi:10.1038/s43588-024-00662-z · Cited by 83 · see citing works
“Differences in imaging modalities, scanners, annotations or patient populations can make models fail when applied to the data of different clinical centers 35 .”
Deep Learning for Cardiovascular Imaging
Wehbe, R. M., Katsaggelos, A. K., Hammond, K. J., Hong, H., Ahmad, F. S., Ouyang, D., Shah, S. J., McCarthy, P. M., & Thomas, J. D. (2023) · JAMA Cardiology · doi:10.1001/jamacardio.2023.3142 · Cited by 71 · see citing works
The Internet of Federated Things (IoFT)
Kontar, R. A., Shi, N., Yue, X., Chung, S., Byon, E., Chowdhury, M., Jin, J. J., Kontar, W., Masoud, N., Nouiehed, M., et al. (2021) · IEEE Access · doi:10.1109/access.2021.3127448 · Cited by 67 · see citing works
“More recently, federated ML has been applied to the development of multi-hospital models in healthcare [261, 266, 126].”
MethodDeep learning-based artificial intelligence applications in prostate MRI: brief summary
Türkbey, B., & Haider, M. A. (2021) · British Journal of Radiology · doi:10.1259/bjr.20210563 · Cited by 66 · see citing works
Self-citationA Review of Medical Federated Learning: Applications in Oncology and Cancer Research
Chowdhury, A., Kassem, H., Padoy, N., Umeton, R., & Karargyris, A. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-08999-2_1 · Cited by 70 · see citing works
“[ 46] performed prostate image segmentation in a federated setting.”
BackgroundMethodProportionally Fair Hospital Collaborations in Federated Learning of Histopathology Images
Hosseini, S. M., Sikaroudi, M., Babaie, M., & Tizhoosh, H. R. (2023) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2023.3234450 · Cited by 59 · see citing works
“The authors in [14] used private prostate MRI images from three institutions and showed the superiority of federated learning in prostate MRI segmentation.”
BackgroundBlika, A., Palmos, S., Doukas, G. S., Lamprou, V., Pelekis, S., Kontoulis, M., Ntanos, C., & Askounis, D. (2024) · IEEE Open Journal of the Communications Society · doi:10.1109/ojcoms.2024.3449563 · Cited by 69 · see citing works
“Reference [138] demonstrate the use of FL at three different institutions on the medical image analysis of entire prostate segmentation, which is a first step towards magnetic resonance imaging (MRI)-guided cancer diagnosis and fusion-guided therapies.”
In the Pursuit of Privacy: The Promises and Predicaments of Federated Learning in Healthcare
Topaloglu, M. Y., Morrell, E. M., Rajendran, S., & Topaloĝlu, Ü. (2021) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2021.746497 · Cited by 50 · see citing works
“Given the large volume and diverse data needed for model training, Federated Learning (FL) approaches may provide a novel opportunity for the future of ML applications (Rajendran et al., 2021; Sarma et al., 2021).”
BackgroundRong, Y., Chen, Q., Fu, Y., Yang, X., Al‐Hallaq, H., Wu, Q., Yuan, L., Xiao, Y., Cai, B., Latifi, K., et al. (2023) · International Journal of Radiation Oncology*Biology*Physics · doi:10.1016/j.ijrobp.2023.10.033 · Cited by 51 · see citing works
Ho, T.-T., Tran, K.-D., & Huang, Y. (2022) · Sensors · doi:10.3390/s22103728 · Cited by 44 · see citing works
Privacy-preserving blockchain-based federated learning for brain tumor segmentation
Kumar, R., Bernard, C. M., Ullah, A., Khan, R. U., Kumar, J., Kulevome, D. K. B., Rao, Y., & Zeng, S. (2024) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2024.108646 · Cited by 45 · see citing works
Gupta, S., Kumar, S., Chang, K., Lu, C., Singh, P., & Kalpathy–Cramer, J. (2023) · Radiographics · doi:10.1148/rg.220107 · Cited by 44 · see citing works
Transparency of medical artificial intelligence systems
Kim, C., Gadgil, S. U., & Lee, S. (2025) · Nature Reviews Bioengineering · doi:10.1038/s44222-025-00363-w · Cited by 58 · see citing works
The Federated Tumor Segmentation (FeTS) ChallengePreprint
Pati, S., Baid, U., Zenk, M., Edwards, B., Sheller, M., Reina, G. A., Foley, P., Gruzdev, A., Martin, J., Albarqouni, S., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2105.05874 · Cited by 102 · see citing works
Choudhury, A., Volmer, L., Martin, F. J., Fijten, R., Wee, L., Dekker, A. L. A. J., & van Soest, J. (2024) · JMIR AI · doi:10.2196/60847 · Cited by 44 · see citing works
“Existing literature has shown that FL models can achieve performance comparable to centrally trained models [38,41,45-47].”
FedPerl: Semi-supervised Peer Learning for Skin Lesion Classification
Bdair, T. M., Navab, N., & Albarqouni, S. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-87199-4_32 · Cited by 46 · see citing works
Prediction across healthcare settings: a case study in predicting emergency department disposition
Barak‐Corren, Y., Chaudhari, P. P., Perniciaro, J. L., Waltzman, M. L., Fine, A. M., & Reis, B. Y. (2021) · npj Digital Medicine · doi:10.1038/s41746-021-00537-x · Cited by 37 · see citing works
Augmenting digital twins with federated learning in medicine
Nagaraj, D., Khandelwal, P., Steyaert, S., & Gevaert, O. (2023) · The Lancet Digital Health · doi:10.1016/s2589-7500(23)00044-4 · Cited by 41 · see citing works
Mazher, M., Razzak, I., Qayyum, A., Tanveer, M., Beier, S., Khan, T. M., & Niederer, S. (2024) · Information Fusion · doi:10.1016/j.inffus.2024.102256 · Cited by 38 · see citing works
Artificial Intelligence for Dermatopathology: Current Trends and the Road Ahead
Chen, S. B., & Novoa, R. A. (2022) · Seminars in Diagnostic Pathology · doi:10.1053/j.semdp.2022.01.003 · Cited by 37 · see citing works
Federated Multi-Organ Segmentation With Inconsistent Labels
Xu, X., Deng, H., Gateño, J., & Yan, P. (2023) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2023.3270140 · Cited by 37 · see citing works
“…barrier of data privacy and realize large-scale DL on isolated data, federated learning (FL) [1], an emerging decentralized learning paradigm, has been adopted in the medical image analysis, solving various clinical problems such as prostate segmentation [2], [3] and COVID-19 diagnosis [4], [5].”
Kumar, R., Waisberg, E., Ong, J., Paladugu, P., Amiri, D., Saintyl, J., Yelamanchi, J., Nahouraii, R., Jagadeesan, R., & Tavakkoli, A. (2024) · Brain Sciences · doi:10.3390/brainsci14121266 · Cited by 51 · see citing works
Advancing breast, lung and prostate cancer research with federated learning. A systematic review
Ankolekar, A., Boie, S. D., Abdollahyan, M., Gadaleta, E., Hasheminasab, S. A., Yang, G., Beauville, C., Δικαίος, Ν., Kastis, G. A., Bussmann, M., et al. (2025) · npj Digital Medicine · doi:10.1038/s41746-025-01591-5 · Cited by 37 · see citing works
“HorizontalFLwasusedmainlyforclassi fi cation tasks 17,27,29,36,38 , followed by detection 20,23,25 , segmentation 25,26 and regression 29 .”
Baek, S., Jung, E.-Y., Pyo, J., Pachepsky, Y., Son, H., & Cho, K. (2022) · Water Research · doi:10.1016/j.watres.2022.118494 · Cited by 27 · see citing works
Federated Partially Supervised Learning With Limited Decentralized Medical Images
Dong, N., Kampffmeyer, M., Voiculescu, I. D., & Xing, E. P. (2022) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2022.3231017 · Cited by 31 · see citing works
“FL is a learning paradigm that aims to utilize decentralized data stored separately in different places and has become a topic of active research in medical image analysis [14], [15], [16], [17], [18], [19].”
BackgroundMulti-institutional PET/CT image segmentation using federated deep transformer learning
Shiri, I., Razeghi, B., Sadr, A. V., Amini, M., Salimi, Y., Ferdowsi, S., Boor, P., Gündüz, D., Voloshynovskiy, S., & Zaidi, H. (2023) · Computer Methods and Programs in Biomedicine · doi:10.1016/j.cmpb.2023.107706 · Cited by 35 · see citing works
Developing and deploying deep learning models in brain magnetic resonance imaging: A review
Aggarwal, K., Jimeno, M. M., Ravi, K. S., González, R. G., & Geethanath, S. (2023) · NMR in Biomedicine · doi:10.1002/nbm.5014 · Cited by 31 · see citing works
“53,54 When dealing with multi-institutional data, protocol harmonization can avoid model bias that may arise from differences in image contrast, intensity, or noise distribution.”
BackgroundPrivacy-Enhancing Technologies in Biomedical Data Science
Cho, H., Froelicher, D., Dokmai, N., Nandi, A., Sadhuka, S., Hong, M. M., & Berger, B. (2024) · Annual Review of Biomedical Data Science · doi:10.1146/annurev-biodatasci-120423-120107 · Cited by 34 · see citing works
An improved federated deep learning for plant leaf disease detection
Hari, P., Singh, M. P., & Singh, A. K. (2024) · Multimedia Tools and Applications · doi:10.1007/s11042-024-18867-9 · Cited by 29 · see citing works
“Sarma et al. [10] demonstrate FL without centralizing or sharing the physical data of multi-institutions.”
Abbas, S., Asif, M. W., Rehman, A., Alharbi, M., Khan, M. A., & Elmitwally, N. S. (2024) · Heliyon · doi:10.1016/j.heliyon.2024.e36743 · Cited by 28 · see citing works
“Integrating federated learning with emerging technologies such as artificial intelligence, deep learning, and computer vision further enhances the precision and personalization of cancer diagnostics [74].”
BackgroundRetractedNovel technical and privacy-preserving technology for artificial intelligence in ophthalmology
Lim, J. S., Hong, M., Lam, W. S. T., Zhang, Z., Teo, Z. L., Liu, Y., Ng, W. Y., Foo, L. L., & Ting, D. S. W. (2022) · Current Opinion in Ophthalmology · doi:10.1097/icu.0000000000000846 · Cited by 30 · see citing works
“…EMR for COVID-19 mortality prediction [140], and genomics data for the construction of type 2 diabetes polygenic risk scores [ outperformedtraditionalsingle-institutionMLarchi-tectures [148,151–156] and may be comparable to models built via central learning (centralized database) [137,138,143–150].”
MethodPreparing Data for Artificial Intelligence in Pathology with Clinical-Grade Performance
Yang, Y., Sun, K., Gao, Y., Wang, K., & Yu, G. (2023) · Diagnostics · doi:10.3390/diagnostics13193115 · Cited by 31 · see citing works
“The National Management Measures for Health Care Big Data Standards, Security and Services (Trial) [50], Health Insurance Portability and Accountability Act (HIPAA) [51] and other related laws should be complied strictly in data collection, storage, usage and disclosure.”
BackgroundRajagopal, A., Redekop, E., Kemisetti, A., Kulkarni, R., Raman, S. S., Sarma, K. V., Magudia, K. A., Arnold, C. W., & Larson, P. E. Z. (2023) · Academic Radiology · doi:10.1016/j.acra.2023.02.012 · Cited by 28 · see citing works
Self-citationApplying Deep Learning-Based Human Motion Recognition System in Sports Competition
Zhang, L. (2022) · Frontiers in Neurorobotics · doi:10.3389/fnbot.2022.860981 · Cited by 26 · see citing works
“Thus, CNN has exerted excellent performance in visual target recognition (Sarma K. V. et al., 2021).”
BackgroundFed2A: Federated Learning Mechanism in Asynchronous and Adaptive Modes
Liu, S., Chen, Q., & You, L. (2022) · Electronics · doi:10.3390/electronics11091393 · Cited by 24 · see citing works
COMMUTE: Communication-efficient transfer learning for multi-site risk prediction
Gu, T., Lee, P. H., & Duan, R. (2022) · Journal of Biomedical Informatics · doi:10.1016/j.jbi.2022.104243 · Cited by 24 · see citing works
Bi, X., Gupta, A., & Yang, M. (2023) · Management Science · doi:10.1287/mnsc.2023.00611 · Cited by 25 · see citing works
Survey of Medical Applications of Federated Learning
Choi, G., Cha, W. C., Lee, S. U., & Shin, S.-Y. (2024) · Healthcare Informatics Research · doi:10.4258/hir.2024.30.1.3 · Cited by 32 · see citing works
“The remaining five studies explored a variety of radiology images, including mammography [24,47], prostate MRI [23], cardiac MRI [52], and pancreatic CT [25].”
BackgroundReal-world federated learning in radiology: hurdles to overcome and benefits to gain
Bujotzek, M., Akünal, Ü., Denner, S., Neher, P., Zenk, M., Frodl, E., Jaiswal, A., Kim, M., Krekiehn, N. R., Nickel, M., et al. (2024) · Journal of the American Medical Informatics Association · doi:10.1093/jamia/ocae259 · Cited by 25 · see citing works
“…legal framework across diverse legal jurisdictions ( ★ ,[13]) →S1: Draft individual legal contracts between the FL initiative and each site ( ★ , [12] ) Step 2: Develop legal framework for sharing of DL models within the FL initiative ( ★ ) - H2: Determine if DL models constitute patient-related…”
BackgroundThe performance of artificial intelligence in prostate magnetic resonance imaging screening
Owida, H. A., Hassan, M. R., Ali, A. M., Alnaimat, F., Sharah, A. A., Abuowaida, S. F. A., & Alshdaifat, N. (2024) · International Journal of Power Electronics and Drive Systems/International Journal of Electrical and Computer Engineering · doi:10.11591/ijece.v14i2.pp2234-2241 · Cited by 24 · see citing works
“In order to speed up the construction of models between institutions and provide higher generalizability in clinical application without actual data exchange, alternatives such as federated learning can be highly effective [54].”
BackgroundA Federated Learning Framework for Brain Tumor Segmentation Without Sharing Patient Data
Zhang, W., Jin, W., Rho, S., Jiang, F., & Yang, C. (2024) · International Journal of Imaging Systems and Technology · doi:10.1002/ima.23147 · Cited by 22 · see citing works
Depression clinical detection model based on social media: a federated deep learning approach
Liu, Y. (2023) · The Journal of Supercomputing · doi:10.1007/s11227-023-05754-7 · Cited by 23 · see citing works
Ali, A., Huang, J., & Jabbar, A. (2025) · IEEE Access · doi:10.1109/access.2025.3562128 · Cited by 21 · see citing works
Insights into artificial intelligence in myopia management: from a data perspective
Zhang, J., & Zou, H. (2023) · Graefe s Archive for Clinical and Experimental Ophthalmology · doi:10.1007/s00417-023-06101-5 · Cited by 18 · see citing works
“Federated learning [108] and swarm learning [109] have emerged as potential methods to cope with privacy problems, providing a decentralized and secure method of data management.”
Lopetuso, L. R., Deleu, S., Puca, P., Abreu, M. T., Armuzzi, A., Barbara, G., Caprioli, F., Ng, S. C., Costello, S. P., Damiani, A., et al. (2025) · Inflammatory Bowel Diseases · doi:10.1093/ibd/izaf013 · Cited by 23 · see citing works
“37,38 Additionally, AI-driven microbiota analysis may enable predictive modeling to identify biomarkers of treatment response, streamline donor matching, and finally improve therapeutic outcomes.”
Federated Learning for Medical Imaging Segmentation via Dynamic Aggregation on Non-IID Data Silos
Yang, L., He, J., Fu, Y., & Luo, Z. (2023) · Electronics · doi:10.3390/electronics12071687 · Cited by 21 · see citing works
Federated Learning for Diabetic Retinopathy Detection in a Multi-center Fundus Screening Network
Matta, S. C., Hassine, M. B., Lecat, C., Borderie, L., Guilcher, A. L., Massin, P., Cochener, B., Lamard, M., & Quellec, G. (2023) · Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) · doi:10.1109/embc40787.2023.10340772 · Cited by 21 · see citing works
Machine learning in metastatic cancer research: Potentials, possibilities, and prospects.
Petinrin, O. O., Saeed, F., Toseef, M., Liu, Z., Basurra, S., Muyide, I. O., Li, X.-T., Lin, Q.-Z., & Wong, K.-C. (2023) · Computational and Structural Biotechnology Journal · doi:10.1016/j.csbj.2023.03.046 · Cited by 19 · see citing works
“It has also shown superior performance compared to data from single institution [172].”
ResultExploring Deep Federated Learning for the Internet of Things: A GDPR-Compliant Architecture
Abbas, Z., Ahmad, S. F., Syed, M. H., Anjum, A., & Rehman, S. (2023) · IEEE Access · doi:10.1109/access.2023.3344029 · Cited by 23 · see citing works
“Real clinical rectal image data were used to effectively illustrate the capability of FL across three academic institutions in [61].”
MethodTrustworthy machine learning in the context of security and privacy
Upreti, R., Lind, P. G., Elmokashfi, A. M., & Yazidi, A. (2024) · International Journal of Information Security · doi:10.1007/s10207-024-00813-3 · Cited by 19 · see citing works
“Similarly, in [105], researchers try to resolve a real-world problem.”
Artificial intelligence in dermatopathology: A systematic review
Lalmalani, R. M., Lim, C. X. Y., & Oh, C. C. (2025) · Clinical and Experimental Dermatology · doi:10.1093/ced/llae361 · Cited by 19 · see citing works
Weimann, K., & Conrad, T. O. F. (2024) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2024.3427787 · Cited by 19 · see citing works
“Federated learning has found a variety of applications in medical research: brain tumor segmentation [20], [21], breast cancer classification [22], whole prostate segmentation [23], lung cancer detection [24] and ECG interpretation [25], [26], [27], [28], [29], [30], [31].”
MethodArtificial Intelligence in Pediatric Electrocardiography: A Comprehensive Review
Leone, D. M., O’Sullivan, D. M., & Jaimes, K. B. (2024) · Children · doi:10.3390/children12010025 · Cited by 23 · see citing works
“However, to ensure equitable outcomes, it is essential that AI model training and validation datasets include diverse populations to prevent the reinforcement or exacerbation of racial, ethnic, and geographic disparities in healthcare.”
Chen, Z., Wang, C., & Wang, F. (2025) · World Journal of Gastroenterology · doi:10.3748/wjg.v31.i24.108021 · Cited by 19 · see citing works
MERGE: A model for multi-input biomedical federated learning
Casella, B., Riviera, W., Aldinucci, M., & Menegaz, G. (2023) · Patterns · doi:10.1016/j.patter.2023.100856 · Cited by 20 · see citing works
“31,32 Due to the high sensitivity of the data, many other works have been performed to explore FL in the biomedical field.”
BackgroundLiu, Y., Huang, J., Chen, J., Chen, W., Pan, Y., & Qiu, J. (2024) · BMC Cancer · doi:10.1186/s12885-024-12456-7 · Cited by 21 · see citing works
“Meanwhile, Federated Learning (FL) [12, 13] has garnered substantial interest within the medical domain as an emerging machine learning technique [14].”
BackgroundPan, J., Fan, Z., Smith, G. E., Guo, Y., Bian, J., & Xu, J. (2025) · Alzheimer's & Dementia · doi:10.1002/alz.70128 · Cited by 19 · see citing works
“57 Given that EHRs include patient treatment data In addition, our demographic analyses uncovered regional varia-tionsthatarecrucialforinterpretingthestudyresults.”
Privacy-preserving collaborative machine learning in biomedical applications
Kim, W., & Seok, J. (2022) · International Conference on Artificial Intelligence in Information and Communication (ICAIIC) · doi:10.1109/icaiic54071.2022.9722703 · Cited by 18 · see citing works
Federated Learning for Medical Imaging: An Updated State of the Art
Mouhni, N., Elkalay, A., Chakraoui, M., Abdali, A., Ammoumou, A., & Amalou, I. (2022) · Ingénierie des systèmes d information · doi:10.18280/isi.270117 · Cited by 17 · see citing works
“Recent works have demonstrated that FL model performance metrics like accuracy, recall and F1-score, are equal or sometimes even better than a single institution trained model [29, 39, 40, 49, 50].”
BackgroundZenk, M., Baid, U., Pati, S., Linardos, A., Edwards, B., Sheller, M., Foley, P., Aristizábal, A., Zimmerer, D., Gruzdev, A. D., et al. (2025) · Nature Communications · doi:10.1038/s41467-025-60466-1 · Cited by 19 · see citing works
Differentially private federated deep learning for multi-site medical image segmentationPreprint
Ziller, A., Usynin, D., Remerscheid, N. W., Knolle, M. A., Braren, R., Rueckert, D., & Kaissis, G. A. (2022) · Research Square · doi:10.21203/rs.3.rs-1478332/v1 · Cited by 18 · see citing works
Federated learning using model projection for multi-center disease diagnosis with non-IID data
Du, J., Li, W., Liu, P., Vong, C., You, Y., Lei, B., & Wang, T. (2024) · Neural Networks · doi:10.1016/j.neunet.2024.106409 · Cited by 15 · see citing works
Federated Learning in Healthcare: From Research to Real-World Deployment
Bakas, S., Li, X., Shah, P., & Roth, H. R. (2026) · Annual Review of Biomedical Engineering · doi:10.1146/annurev-bioeng-080125-041414 · Cited by 19 · see citing works
Self-citationWang, D., Wang, D., Hu, Q., et al. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.123869 · Cited by 15 · see citing works
Provenzano, D., Melnyk, O., Imtiaz, D., McSweeney, B., Nemirovsky, D., Wynne, M., Whalen, M. J., Rao, Y. J., Loew, M. H., & Haji-Momenian, S. (2023) · Applied Sciences · doi:10.3390/app13021088 · Cited by 13 · see citing works
Machine Learning for Hepatocellular Carcinoma Segmentation at MRI:RadiologyIn Training
Raman, A. G., Jones, C. K., & Weiss, C. R. (2022) · Radiology · doi:10.1148/radiol.212386 · Cited by 12 · see citing works
Self-citationContinual learning framework for a multicenter study with an application to electrocardiogram
Kim, J., Lim, M. H., Kim, K., & Yoon, H. (2024) · BMC Medical Informatics and Decision Making · doi:10.1186/s12911-024-02464-9 · Cited by 13 · see citing works
“In federated learning, models are distributed to be trained, aggregated periodically, and distributed again [5].”
Multimodal Wearable Biosensing Meets Multidomain AI: A Pathway to Decentralized Healthcare
Liu, C., Fan, H., Kim, M., Zhou, T., Yang, P., Zhao, L., Wang, Y., Che, Z., Liu, C., Li, B., et al. (2026) · Advanced Science · doi:10.1002/advs.202522900 · Cited by 20 · see citing works
Artificial intelligence in nephrology: predicting CKD progression and personalizing treatment
Yuan, S., Guo, L., & Xu, F. (2025) · International Urology and Nephrology · doi:10.1007/s11255-025-04878-4 · Cited by 12 · see citing works
Jaswanth, P. K., Chowdary, P. Y., & Ramprasad, M. V. S. (2023) · Measurement Sensors · doi:10.1016/j.measen.2023.100868 · Cited by 10 · see citing works
Vo, V. T.-T., Shin, T.-H., Yang, H.-J., Kang, S., & Kim, S.-H. (2024) · Computer Methods and Programs in Biomedicine · doi:10.1016/j.cmpb.2024.108104 · Cited by 10 · see citing works
Towards practical federated learning and evaluation for medical prediction models
Kazlouski, A., Perez, I. M., Noor, F. A., Högerman, M., Ettala, O. O., Pahikkala, T., & Airola, A. (2025) · International Journal of Medical Informatics · doi:10.1016/j.ijmedinf.2025.106046 · Cited by 11 · see citing works
Tsui, J. M. G., Kehayias, C. E., Leeman, J. E., Nguyen, P. L., Peng, L. C., Yang, D. D., Moningi, S., Martin, N. E., Orio, P. F., D'Amico, A. V., et al. (2023) · International Journal of Radiation Oncology*Biology*Physics · doi:10.1016/j.ijrobp.2023.07.029 · Cited by 9 · see citing works
Sakhawat, A. R., Fatima, A., Abbas, S., Ahmad, M., & Khan, M. A. (2024) · Advances in business information systems and analytics book series · doi:10.4018/979-8-3693-0839-4.ch012 · Cited by 10 · see citing works
Escribà-Montagut, X., Marcon, Y., Anguita‐Ruiz, A., Avraam, D., Urquiza, J., Morgan, A. S., Wilson, B., Burton, P. R., & González, J. R. (2024) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1012626 · Cited by 10 · see citing works
Harnessing clinical annotations to improve deep learning performance in prostate segmentation
Sarma, K. V., Raman, A. G., Dhinagar, N. J., Priester, A. M., Harmon, S. A., Sanford, T., Mehralivand, S., Türkbey, B., Marks, L. S., Raman, S. S., et al. (2021) · PLoS ONE · doi:10.1371/journal.pone.0253829 · Cited by 10 · see citing works
Self-citationWang, X., He, J., Ding, G., Tang, Y., & Wang, Q. (2026) · Journal of Multidisciplinary Healthcare · doi:10.2147/jmdh.s572089 · Cited by 11 · see citing works
“119,120 In pharmacogenomics, federated AI has already demonstrated the feasibility of discovering chemotherapy-response biomarkers across hospitals without centra-lizing raw patient data.”
- Authors
- Sarma, K. V., Harmon, S., Sanford, T., Roth, H. R., Xu, Z., Tetreault, J., Xu, D., Flores, M. G., Raman, A. G., Kulkarni, R., Wood, B. J., Choyke, P. L., Priester, A. M., Marks, L. S., Raman, S. S., Enzmann, D., Turkbey, B., Speier, W., & Arnold, C. W.
- Affiliations
- University of California, Los Angeles (US)
- National Institutes of Health (US)
- SUNY Upstate Medical University (US)
- Nvidia (United States) (US)
- National Cancer Institute, National Institutes of Health, Bethesda, Maryland, USA
- Corresponding author
- Arnold, Corey W
- ORCID
- Venue
- Journal of the American Medical Informatics Association
- Publisher
- Oxford University Press (OUP)
- Published
- 2021-02-04
- Volume
- 28
- Issue
- 6
- Pages
- 1259-1264
- ISSN
- 1067-5027 (ISSN-L)
- 1527-974X
- PMC
- PMC8200268
- OpenAlex
- W3128304793
- Semantic Scholar
- ef5abdbf50baa6f719338ffa8a4fe1e41fed0352
- Cited by
- 242
- References
- 25
- Citations by year
- 2026: 27 · 2025: 51 · 2024: 54 · 2023: 48 · 2022: 34 · 2021: 18
- FWCI
- 9.90
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- Journal of the American Medical Informatics Association (publishedVersion, cc-by)
- PubMed (publishedVersion)
- eScholarship (California Digital Library) (submittedVersion, cc-by)
- PubMed Central (submittedVersion, cc-by)
- eScholarship (California Digital Library) (submittedVersion)
- Funders
- NIH NCI (F30CA210329, R21CA220352, P50CA092131, ZIDBC011242, Z1ACL040015, HHSN261200800001E)
- NIH NIGMS (GM08042)
- NIH
- AMA Foundation
- NVIDIA Corporation Academic Hardware Grant
- NIH Center for Interventional Oncology
- Intramural Research Program of the NIH, and a cooperative research and development agreement (CRADA) between NIH and nVIDIA
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Privacy-Preserving Technologies in Data
- Health Sciences › Medicine › Health Informatics › Artificial Intelligence in Healthcare and Education
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Keywords
- federated learning, data privacy, data sharing, model generalization, clinical data
- MeSH terms
- Deep Learning, Humans, Privacy, Information Dissemination
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-07.
Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-07.
Cited sources are checked by gpt-6-sol.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
