Federated learning generator loss.

<div><p>Deep learning models for diagnostic applications require large amounts of sensitive patient data, raising privacy concerns under centralized training paradigms. We propose FedGAN, a federated learning framework for synthetic medical image generation that combines Generative Adver...

Full description

Saved in:
Bibliographic Details
Main Author: Hassan Kamran (21782056) (author)
Other Authors: Syed Jawad Hussain (18106616) (author), Sohaib Latif (20314613) (author), Imtiaz Ali Soomro (21609078) (author), Mrim M. Alnfiai (19226195) (author), Nouf Nawar Alotaibi (21782059) (author)
Published: 2025
Subjects:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1852018207943032832
author Hassan Kamran (21782056)
author2 Syed Jawad Hussain (18106616)
Sohaib Latif (20314613)
Imtiaz Ali Soomro (21609078)
Mrim M. Alnfiai (19226195)
Nouf Nawar Alotaibi (21782059)
author2_role author
author
author
author
author
author_facet Hassan Kamran (21782056)
Syed Jawad Hussain (18106616)
Sohaib Latif (20314613)
Imtiaz Ali Soomro (21609078)
Mrim M. Alnfiai (19226195)
Nouf Nawar Alotaibi (21782059)
author_role author
dc.creator.none.fl_str_mv Hassan Kamran (21782056)
Syed Jawad Hussain (18106616)
Sohaib Latif (20314613)
Imtiaz Ali Soomro (21609078)
Mrim M. Alnfiai (19226195)
Nouf Nawar Alotaibi (21782059)
dc.date.none.fl_str_mv 2025-07-24T17:48:00Z
dc.identifier.none.fl_str_mv 10.1371/journal.pone.0326579.g017
dc.relation.none.fl_str_mv https://figshare.com/articles/figure/Federated_learning_generator_loss_/29640277
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Microbiology
Cell Biology
Genetics
Molecular Biology
Biotechnology
Science Policy
Infectious Diseases
Biological Sciences not elsewhere classified
Chemical Sciences not elsewhere classified
Information Systems not elsewhere classified
sensitive patient data
gan &# 8217
centralized training paradigms
abdominal ct scans
silo federated learning
federated learning framework
raising privacy concerns
fedgan generates high
centralized discriminator ).
federated averaging
privacy challenges
realism score
propose fedgan
medical ai
generator via
fedgan achieves
experiments demonstrate
approach pretrains
dc.title.none.fl_str_mv Federated learning generator loss.
dc.type.none.fl_str_mv Image
Figure
info:eu-repo/semantics/publishedVersion
image
description <div><p>Deep learning models for diagnostic applications require large amounts of sensitive patient data, raising privacy concerns under centralized training paradigms. We propose FedGAN, a federated learning framework for synthetic medical image generation that combines Generative Adversarial Networks (GANs) with cross-silo federated learning. Our approach pretrains a DCGAN on abdominal CT scans and fine-tunes it collaboratively across clinical silos using diabetic retinopathy datasets. By federating the GAN’s discriminator and generator via the Federated Averaging (FedAvg) algorithm, FedGAN generates high-quality synthetic retinal images while complying with HIPAA and GDPR. Experiments demonstrate that FedGAN achieves a realism score of 0.43 (measured by a centralized discriminator). This work bridges data scarcity and privacy challenges in medical AI, enabling secure collaboration across institutions.</p></div>
eu_rights_str_mv openAccess
id Manara_e2ef9805e7c36c20edf4b13b4d48dbce
identifier_str_mv 10.1371/journal.pone.0326579.g017
network_acronym_str Manara
network_name_str ManaraRepo
oai_identifier_str oai:figshare.com:article/29640277
publishDate 2025
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling Federated learning generator loss.Hassan Kamran (21782056)Syed Jawad Hussain (18106616)Sohaib Latif (20314613)Imtiaz Ali Soomro (21609078)Mrim M. Alnfiai (19226195)Nouf Nawar Alotaibi (21782059)MicrobiologyCell BiologyGeneticsMolecular BiologyBiotechnologyScience PolicyInfectious DiseasesBiological Sciences not elsewhere classifiedChemical Sciences not elsewhere classifiedInformation Systems not elsewhere classifiedsensitive patient datagan &# 8217centralized training paradigmsabdominal ct scanssilo federated learningfederated learning frameworkraising privacy concernsfedgan generates highcentralized discriminator ).federated averagingprivacy challengesrealism scorepropose fedganmedical aigenerator viafedgan achievesexperiments demonstrateapproach pretrains<div><p>Deep learning models for diagnostic applications require large amounts of sensitive patient data, raising privacy concerns under centralized training paradigms. We propose FedGAN, a federated learning framework for synthetic medical image generation that combines Generative Adversarial Networks (GANs) with cross-silo federated learning. Our approach pretrains a DCGAN on abdominal CT scans and fine-tunes it collaboratively across clinical silos using diabetic retinopathy datasets. By federating the GAN’s discriminator and generator via the Federated Averaging (FedAvg) algorithm, FedGAN generates high-quality synthetic retinal images while complying with HIPAA and GDPR. Experiments demonstrate that FedGAN achieves a realism score of 0.43 (measured by a centralized discriminator). This work bridges data scarcity and privacy challenges in medical AI, enabling secure collaboration across institutions.</p></div>2025-07-24T17:48:00ZImageFigureinfo:eu-repo/semantics/publishedVersionimage10.1371/journal.pone.0326579.g017https://figshare.com/articles/figure/Federated_learning_generator_loss_/29640277CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/296402772025-07-24T17:48:00Z
spellingShingle Federated learning generator loss.
Hassan Kamran (21782056)
Microbiology
Cell Biology
Genetics
Molecular Biology
Biotechnology
Science Policy
Infectious Diseases
Biological Sciences not elsewhere classified
Chemical Sciences not elsewhere classified
Information Systems not elsewhere classified
sensitive patient data
gan &# 8217
centralized training paradigms
abdominal ct scans
silo federated learning
federated learning framework
raising privacy concerns
fedgan generates high
centralized discriminator ).
federated averaging
privacy challenges
realism score
propose fedgan
medical ai
generator via
fedgan achieves
experiments demonstrate
approach pretrains
status_str publishedVersion
title Federated learning generator loss.
title_full Federated learning generator loss.
title_fullStr Federated learning generator loss.
title_full_unstemmed Federated learning generator loss.
title_short Federated learning generator loss.
title_sort Federated learning generator loss.
topic Microbiology
Cell Biology
Genetics
Molecular Biology
Biotechnology
Science Policy
Infectious Diseases
Biological Sciences not elsewhere classified
Chemical Sciences not elsewhere classified
Information Systems not elsewhere classified
sensitive patient data
gan &# 8217
centralized training paradigms
abdominal ct scans
silo federated learning
federated learning framework
raising privacy concerns
fedgan generates high
centralized discriminator ).
federated averaging
privacy challenges
realism score
propose fedgan
medical ai
generator via
fedgan achieves
experiments demonstrate
approach pretrains