A comprehensive survey on federated learning techniques for healthcare informatics

Classified as probably not a substantial claim.

FL allows multiple parties to collaborate on model training without exchanging or centralizing datasets.

“FL makes it possible for several parties to work together without exchanging or centralizing datasets.”

• Also stated in: § 3.4. Main Contributions of the Survey · plain

  “FL facilitates communication between various parties without the need to communicate or centralize data, thereby resolving issues with critical health information.”

May combine two claims.

FL reduces data security and privacy concerns by maintaining local data stores, unlike centralized ML which requires datasets to reside on a single server.

“In contrast to traditional centralized ML methods, which require datasets to reside on a single server, FL minimizes security and privacy concerns by keeping local data stores [12, 13].”

• Also stated in: § 3.3. Motivations of Using FL for Healthcare · plain

  “FL reduces data security and privacy concerns by maintaining local data stores, as opposed to centralized ML techniques, which require datasets to reside on one server [51].”

• cites [12] Khan R. U. Zhang X. Alazab M. Ku- mar R. An improved convolutional neural network model for intrusion detection in networks Proceedings of the 2019 Cybersecurity and cyberforensics confer- ence (CCC) May 2019 Melbourne, VIC, Australia IEEE 74 77 10.1109/ccc.2019.000-6 2-s2.0-85073870196 sourceNot checked yet

  Not checked in the free run.

• cites [13] Alazab M. Broadhurst R. Spam and crimi- nal activity Trends and issues in crime and crim- inal justice 2016 526 1 20Not checked yet

  Not checked in the free run.

• cites [51] Song J. Wang W. Gadekallu T. R. Cao J. Liu Y. Eppda: An efficient privacy-preserving data aggregation federated learning scheme IEEE Transactions on Network Science and En- gineering 2022Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

FL offers great assurances on analytics of health data.

“FL offers great assurances on analytics of health data.”

Probably not a substantial claim: may not pass S3 (load-bearing).

FedHealth creates capable ML models based on FL and homomorphic encryption to protect users' privacy.

“FedHealth software creates capable ML models based on FL and homomorphic encryption to protect users' privacy.”

Probably not a substantial claim: may not pass S3 (load-bearing).

FL is a paradigm that aims to collaborate on data management and privacy issues by using evaluation metrics without transferring samples.

“FL is a paradigm that aims to collaborate on data management and privacy issues by using evaluation metrics without transferring samples [47–49].”

• cites [47] McMahan B. Moore E. Ramage D. Hamp- son S. y Arcas B. A. Communication-efficient learning of deep networks from decentralized data Artificial Intelligence and Statistics 2017 PMLR 1273 1282Not checked yet

  Not checked in the free run.

• cites [48] Li T. Sahu A. K. Talwalkar A. Smith V. Federated learning: challenges, methods, and future directions IEEE Signal Processing Magazine 2020 37 3 50 60 10.1109/msp.2020.2975749 sourceNot checked yet

  Not checked in the free run.

• cites [49] Arikumar K. Prathiba S. B. Alazab M. Fl-pmi: federated learning-based per- son movement identification through wearable devices in smart healthcare systems Sensors 2022 22 4 p. 1377 10.3390/s22041377 PMC8962969 35214282 sourceNot checked yet

  Not checked in the free run.

May combine two claims.

Health records in the public sector are often scattered and confidential, making it difficult to attain reliable outcomes.

“Health records in the public sector are often scattered and confidential, making it difficult to attain reliable outcomes.”

• cites [6] Chen M. Mathews R. Ouyang T. Bea- ufays F. Federated learning of out-of-vocabulary words 2019 arXiv preprint arXiv:1903.10635Not checked yet

  Not checked in the free run.

• cites [7] Hard A. Rao K. Mathews R. Federated learning for mobile keyboard prediction 2018 arXiv preprint arXiv:1811.03604Not checked yet

  Not checked in the free run.

• cites [8] McMahan H. B. Ramage D. Talwar K. Zhang L. Learning differentially pri- vate recurrent language models 2017 arXiv preprint arXiv:1710.06963Not checked yet

  Not checked in the free run.

• cites [9] Ramaswamy S. Mathews R. Rao K. Beaufays F. Federated learning for emoji Pre- diction in a mobile keyboard 2019 arXiv preprint arXiv:1906.04329Not checked yet

  Not checked in the free run.

• cites [10] Yang T. Andrew G. Eichner H. Applied federated learning: improv- ing google keyboard query suggestions 2018 arXiv preprint arXiv:1812.02903Not checked yet

  Not checked in the free run.

Classified as probably not a substantial claim.

FL allows findings to be obtained collectively between entities without sharing patient data, such as in the form of a universal model.

“In the context of electronic medicine, it suggests FL allows findings to be obtained collectively between entities without sharing patient data, such as in the form of a universal model.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

In healthcare, FL is often used to study tasks such as client identity, client knowledge, diagnostics, identifying possible hospitalizations, detecting death rates, and intensive care unit admission periods.

“In healthcare, FL is often used to study a range of tasks, like client identity, client knowledge, diagnostics, identifying possible hospitalizations, detecting death rates, intensive care unit admission periods, etc.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

ML algorithms require vast, comprehensive datasets, and FL provides controlled, indirect access while protecting patient privacy.

“ML algorithms require vast, comprehensive datasets, and FL provides controlled, indirect access while protecting patient privacy.”

Probably not a substantial claim: may not pass S3 (load-bearing).

FedHealth uses transfer learning techniques to ensure that each organization has a uniquely tailored model after improving its cloud model.

“FedHealth uses transfer learning techniques to ensure that each organization has a uniquely tailored model after improving its cloud model.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

FedHealth can update the cloud model and user model simultaneously as soon as new data is available.

“FedHealth can update the cloud model and user model simultaneously as soon as new data is available.”

Probably not a substantial claim: may not pass S3 (load-bearing).

FedHealth is an FL framework being developed by Chinese researchers to minimize the risk of security breaches.

“An FL framework called FedHealth is being developed by Chinese researchers in order to minimize the risk of security breaches.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

FL brings together all healthcare institutions, allowing them to share experiences and ensuring their privacy.

“FL brings together all the healthcare institutions, allowing them to share experiences and ensuring their privacy.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

FTL employs cryptography and estimation to verify that confidentiality is successfully preserved to prevent potentially exposed client data.

“FTL employs cryptography and estimation to verify that confidentiality is successfully preserved to prevent potentially exposed client data.”

Probably not a substantial claim: may not pass S4 (based).

FL allows data isolation to be solved by combining user data to build ML models.

“FL allows data isolation to be solved by combining user data to build ML models.”

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

Large, diverse datasets are essential for the success of data-driven applications.

“Large, diverse datasets are essential for the success of data-driven applications.”

May combine two claims.

FL solves medical data problems by allowing group work by centralizing information, and it is already being utilized in digital applications across the health field.

“FL solves these problems by allowing group work by centralizing information, and it is already being utilized in digital applications across the health field [31].”

• cites [31] Rieke N. Hancox J. Li W. The future of digital health with federated learning NPJ digital medicine 2020 3 1 1 7 10.1038/s41746-020-00323-1 33015372 PMC7490367 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S4 (based).

It is difficult to get medical datasets.

“However, it is difficult to get medical datasets.”

Probably not a substantial claim: may not pass S3 (load-bearing).

FL is proposed as a promising approach to COVID-19 detection.

“FL is proposed as a promising approach to COVID-19 detection.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based). May combine two claims.

FL could lead to new business opportunities and improved healthcare worldwide.

“In addition, this could lead to new business opportunities and improved healthcare worldwide.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

FL is developing and enabling applications in healthcare such as medical image processing, IoT-based smart healthcare applications, and outbreak prediction.

“FL is next developing and enabling applications in healthcare such as medical image processing, IoT-based smart healthcare applications, and outbreak prediction.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

DL-based MRI reconstruction methods necessarily involve massive amounts of data, which are difficult to generate and distribute due to high acquisition costs and medical data privacy regulations.

“However, these methods necessarily involve massive amounts of data, which are difficult to generate and distribute due to high acquisition costs and medical data privacy regulations.”

Classified as probably not a substantial claim.

Many challenges and issues are imposed in FL for healthcare in terms of confidentiality and safety, data heterogeneity, traceability and accountability, and system architecture.

“Despite several research and development activities, many challenges and issues are imposed in FL for healthcare. We present these difficulties in terms of confidentiality and safety, data heterogeneity, traceability and accountability, and system architecture.”

Probably not a substantial claim: may not pass S4 (based).

Construction of good models without biases using small datasets is very difficult as it takes time, effort, and cost to collect, curate, and maintain high-quality data from a diverse population.

“The construction of good models without biases using small datasets is very difficult as it takes a lot of time, effort, and cost to collect, curate, and maintain high-quality data from a diverse population.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

FL brings together multiple hospitals to develop collaborative healthcare settings that speed up diagnostic testing of victims while maintaining their personal rights.

“FL brings together multiple hospitals to develop collaborative healthcare settings that speed up diagnostic testing of victims while maintaining their personal rights.”

Probably not a substantial claim: may not pass S4 (based). May combine two claims.

FL provides AI developers with access to bigger and diversified data packages, which better portray current patients.

“FL provides AI developers with access to bigger and diversified data packages, which better portray current patients.”

May combine two claims.

With access to electronic health records, FL can help to find clinically similar patients and predict hospitalizations due to cardiac events, mortality and ICU stay time.

“For example, with access to electronic health records, FL can help to find clinically similar patients and predict hospitalizations due to cardiac events, mortality and ICU stay time [52].”

• cites [52] Huang L. Shea A. L. Qian H. Masurkar A. Deng H. Liu D. Patient clustering improves efficiency of federated machine learning to pre- dict mortality and hospital stay time using dis- tributed electronic medical records Journal of Biomedical Informatics 2019 99 10.1016/j.jbi.2019.103291 2-s2.0-85073123575 31560949 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

FL can have an enormous influence on a variety of stakeholders like clinicians, patients, hospitals, medical researchers, and healthcare providers.

“FL can have an enormous influence on a variety of stakeholders like clinicians, patients, hospitals, medical researchers, and healthcare providers.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

The healthcare sector is one of the most attractive applications of IoT.

“The healthcare sector is one of the most attractive applications of IoT, as depicted in Figure 7.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Medical information requires a degree of insensitivity and is regulated by medical rules like HIPAA.

“Medical information requires a degree of insensitivity and is regulated by medical rules like the Health Insurance Portability and Accountability Act (HIPAA) [64].”

• cites [64] Drolet B. C. Marwaha J. S. Hyatt B. Blazar P. E. Lifchez S. D. Electronic communi- cation of protected health information: privacy, security, and hipaa compliance The Journal of Hand Surgery 2017 42 6 411 416 10.1016/j.jhsa.2017.03.023 2-s2.0-85044110572 28578767 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

AI-based healthcare solutions will be able to scale globally at an unprecedented level as a result of FL.

“As a result, AI-based healthcare solutions will be able to scale globally at an unprecedented level.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

DL-based techniques have been discovered to perform better in MRI reconstruction.

“In recent years, DL-based techniques have been discovered to perform better in MRI reconstruction.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing). May combine two claims.

FL is more potential in analyzing medical images and also in protecting the patient's privacy.

“FL is more potential in analyzing the medical images and also in protecting the patient's privacy.”

May combine two claims.

Through FL, large datasets from multiple hospitals can be readily accessed by an individual hospital without centralizing the data into one place, addressing data access rights and access to heterogeneous data.

“Through FL, large datasets from multiple hospitals can be readily accessed by an individual hospital without centralizing the data into one place. As a result of this practice, critical issues such as data access rights and access to heterogeneous data are addressed.”

Probably not a substantial claim: may not pass S3 (load-bearing).

The use of AI-based approaches to learn health data has been adopted widely in smart healthcare, such as the use of intelligent imaging to detect disease.

“The use of AI-based approaches to learn health data has been adopted widely in smart healthcare, such as the use of intelligent imaging to detect disease [63].”

• cites [63] Shickel B. Tighe P. J. Bihorac A. Rashidi P. Deep ehr: a survey of recent advances in deep learning techniques for electronic health record (ehr) analysis IEEE journal of biomedical and health informatics 2017 22 5 1589 1604 29989977 PMC6043423 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

FL makes it possible for several workers to learn a strong and generic ML model by providing information so that key issues like user privacy, information security, authentication rights, and exposure to large datasets are addressed.

“FL makes it possible for several workers to learn a strong and generic ML model by providing information so that key issues like user privacy, information security, authentication rights, and exposure to large datasets are addressed.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Google first suggested federated machine learning in 2016, training ML models on distributed mobile phones.

“Google first suggested federated machine learning [Kone Diecn'y et al., 2016], in which it trained ML models on distributed mobile phones.”

• cites [Kone Diecn'y et al., 2016]Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

FL captures a wide range of data variables and analyzes patients based on any of their demographic characteristics.

“FL captures a wide range of data variables and analyzes patients based on any of their demographic characteristics.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The FL algorithm is used to describe and identify physician-related persons and to estimate their hospitalization, deaths, and survival rates based on electronic health records.

“The FL algorithm is used to describe and identify physician-related persons and to estimate their hospitalization, deaths, and survival rates based on electronic health records (EHR) [54].”

• cites [54] Brisimi T. S. Chen R. Mela T. Olshevsky A. Paschalidis I. C. Shi W. Federated learn- ing of predictive models from federated electronic health records International Journal of Medical Informatics 2018 112 59 67 10.1016/j.ijmedinf.2018.01.007 2-s2.0-85041413697 29500022 PMC5836813 sourceNot checked yet

  Not checked in the free run.

May combine two claims.

FL enables efficient and accurate heart disease diagnosis.

“FL enables efficient and accurate heart disease diagnosis.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

Hospitals can train their own AI models by uploading the data to a global aggregator.

“Hospitals can train their own AI models by uploading the data to a global aggregator.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

FL is a concept developed to construct smart and confidential IoT devices.

“FL is a concept developed to construct smart and confidential IoT devices.”

Probably not a substantial claim: may not pass S2 (checkable).

FL is a potential concept for safe, reliable, and impartial models of data.

“FL is a potential concept for safe, reliable, and impartial models of data.”

Classified as probably not a substantial claim.

FL is capable of resolving data isolation issues through network model training in the context of privacy preservation.

“FL is capable of resolving data isolation issues through network model training in the context of privacy preservation.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

According to studies, the incidence of AKI ranges between 46 percent and 71 percent; however, most deaths occur within the AKI sample.

“According to studies, the incidence of AKI ranges between 46 percent and 71 percent; however, most deaths occur within the AKI sample [87, 88].”

• cites [87] Zahid U. Ramachandran P. Spitale- witz S. Acute kidney injury in COVID-19 patients: an inner city hospital experience and policy implications American Journal of Nephrology 2020 51 10 786 796 10.1159/000511160 33011717 PMC7573899 sourceNot checked yet

  Not checked in the free run.

• cites [88] Hirsch J. S. Ng J. H. Ross D. W. Acute kid- ney injury in patients hospitalized with COVID-19 Kidney International 2020 98 1 209 218 10.1016/j.kint.2020.05.006 32416116 PMC7229463 sourceNot checked yet

  Not checked in the free run.

May combine two claims.

FL delivers similar accuracy in predicting ADR without sacrificing data confidentiality in comparison with centralized AI approaches.

“FL delivers similar accuracy in predicting ADR without sacrificing data confidentiality in comparison with centralized AI approaches.”

• cites [70] Choudhury O. Park Y. Salonidis T. Gkoulalas-Divanis A. Sylla I. Pre- dicting adverse drug reactions on distributed health data using federated learning AMIA Annual symposium proceedings 2019 2019 313 32308824 PMC7153050Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Diagnostic methods that judiciously help individuals at greater risk of contracting COVID-19 can be beneficial during an outbreak where facilities can be restricted.

“Diagnostic methods that judiciously help individuals at greater risk of contracting COVID-19 can be beneficial during an outbreak where facilities can be restricted [89].”

• cites [89] Goldfarb D. S. Benstein J. A. Zhdanova O. Impending shortages of kidney replacement therapy for COVID-19 pa- tients Clinical Journal of the American Society of Nephrology 2020 15 6 880 882 10.2215/cjn.05180420 32345750 PMC7274293 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

FL can ensure proper healthcare coordination for better provision of medical facilities through its centralized and secure nature.

“In order to contribute, FL can ensure proper healthcare coordination for better provision of medical facilities through its centralized and secure nature.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

FL allows the sharing of EHRs without their sharing due to its dynamic approach to education.

“FL allows the sharing of EHRs without their sharing due to its dynamic approach to education.”

May combine two claims.

EHR information can be used to detect rare ADRs at a single site and to predict more severe ADRs in the long term.

“It can be used to detect rare ADRs at a single site for rare ADRs and to predict more severe ADRs in the long term [70].”

• cites [70] Choudhury O. Park Y. Salonidis T. Gkoulalas-Divanis A. Sylla I. Pre- dicting adverse drug reactions on distributed health data using federated learning AMIA Annual symposium proceedings 2019 2019 313 32308824 PMC7153050Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

Traditional AI systems relying on some kind of master database software are not really suitable for contemporary health.

“Traditional AI systems relying on some kind of master database software are not really suitable for contemporary health.”

Probably not a substantial claim: may not pass S3 (load-bearing).

Yuan et al. present a cooperative framework for healthcare that leverages FL to enable medical IoT devices.

“Yuan et al. [73] present a cooperative framework for healthcare that leverages FL to enable medical IoT devices.”

• cites [73] Yuan B. Ge S. Xing W. A federated learn- ing framework for healthcare iot devices 2020 arXiv preprint arXiv:2005.05083Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Personalization training on edge computers was decided to be carried out to reduce heterogeneity and achieve high-quality individual models.

“Personalization training on edge computers was decided to be carried out, in this case, to reduce heterogeneity and achieve high-quality individual models.”

Probably not a substantial claim: may not pass S3 (load-bearing).

A number of clinical symptoms are observed in COVID-19 patients hospitalized with Acute Kidney Injury (AKI).

“A number of clinical symptoms are observed in COVID- 19 patients hospitalized with Acute Kidney Injury (AKI) [83–86].”

• cites [83] Wu Z. McGoogan J. M. Characteristics of and important lessons from the coronavirus dis- ease 2019 (COVID-19) outbreak in China: summary of a report of 72 314 cases from the Chinese cen- ter for disease control and prevention JAMA 2020 323 13 1239 1242 10.1001/jama.2020.2648 32091533 sourceNot checked yet

  Not checked in the free run.

• cites [84] Chung M. Bernheim A. Mei X. Ct imaging features of 2019 novel coronavirus (2019-ncov) Radiology 2020 295 1 202 207 10.1148/radiol.2020200230 32017661 PMC7194022 sourceNot checked yet

  Not checked in the free run.

• cites [85] Lala A. Johnson K. W. Januzzi J. L. Prevalence and im- pact of myocardial injury in patients hospitalized with COVID-19 infection Journal of the American College of Cardiology 2020 76 5 533 546 10.1016/j.jacc.2020.06.007 32517963 PMC7279721 sourceNot checked yet

  Not checked in the free run.

• cites [86] Chan L. Chaudhary K. Saha A. Aki in hospitalized patients with COVID-19 Journal of the American Society of Nephrology 2021 32 1 151 160 10.1681/asn.2020050615 32883700 PMC7894657 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

COVID-19 infection has been confirmed in China since December 2019, and the outbreak began spreading to China and several other countries across the globe.

“COVID-19 infection has been confirmed in China since December 2019. After that, the outbreak began spreading to China and several other countries across the globe [79].”

• cites [79] Hui D. S. Azhar E. I. Madani T. A. The continuing 2019- ncov epidemic threat of novel coronaviruses to global health—the latest 2019 novel coronavirus outbreak in wuhan, China International Journal of Infectious Diseases 2020 91 264 266 10.1016/j.ijid.2020.01.009 31953166 PMC7128332 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

There are currently a number of different ways to detect COVID-19, whereas computed tomography and X-rays remain the major surgical techniques.

“There are currently a number of different ways to detect COVID-19, whereas computed tomography and X-rays remain the major surgical techniques [80–82].”

• cites [80] Kanne J. P. Little B. P. Chung J. H. Elicker B. M. Ketai L. H. Essentials for Radiol- Ogists on COVID-19: An Update—Radiology Scien- Tific Expert Panel 2020 PMC7233379 32105562Not checked yet

  Not checked in the free run.

• cites [81] Xie X. Zhong Z. Zhao W. Zheng C. Wang F. Liu J. Chest ct for typical coronavirus dis- ease 2019 (COVID-19) pneumonia: relationship to negative rt-pcr testing Radiology 2020 296 2 E41 E45 10.1148/radiol.2020200343 32049601 PMC7233363 sourceNot checked yet

  Not checked in the free run.

• cites [82] Zu Z. Y. Jiang M. D. Xu P. P. Chen W. Lu G. M. Zhang L. J. Coronavirus dis- ease 2019 (COVID-19): a perspective from China Radiology 2020 296 2 E15 E25 10.1148/radiol.2020200490 32083985 PMC7233368 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

An innovative method is the use of chain-directed synchronous stochastic gradient descent to minimize human latency between FL clients and servers in personal mobile sensing applications.

“An innovative method [77] is the use of chain-directed synchronous stochastic gradient descent to minimize human latency between FL clients and servers in personal mobile sensing applications.”

• cites [77] Zhang Y. Gu T. Zhang X. Mdldroid: a chainsgd-reduce approach to mobile deep learn- ing for personal mobile sensing Proceedings of the 2020 19th ACM/IEEE International Conference on Infor- mation Processing in Sensor Networks (IPSN) 2020 IEEE 73 84 10.1109/ipsn48710.2020.00-45 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

An FL-based student-centered architecture is responsible for an EHR system with different clinics and a database server.

“An FL-based student-centered architecture is responsible for an EHR system with different clinics as well as a database server [66].”

• cites [66] Hao M. Li H. Xu G. Liu Z. Chen Z. Privacy-aware and resource-saving collaborative learning for healthcare in cloud computing Proceedings of the ICC 2020-2020 IEEE International Conference on Communications (ICC) 2020 Dublin, Ireland IEEE 1 6 10.1109/icc40277.2020.9148979 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

The FedHealth framework was developed for the purpose of stimulating the federation of portable devices using FL technology.

“For the purpose of stimulating the federation of portable devices using FL technology, the FedHealth framework [75] was developed.”

• cites [75] Chen Y. Qin X. Wang J. Yu C. Gao W. Fedhealth: A federated transfer learning frame- work for wearable healthcare IEEE Intelligent Systems 2020 35 4 83 93 10.1109/mis.2020.2988604 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Information collected by a number of healthcare institutions using different IoT devices may be collected using Fed-Health's surgical instruments to enable the development of a powerful AI method that can be used to identify people's behavior and to protect ciphertext data.

“Information collected by a number of healthcare institutions using different IoT devices may be collected using Fed- Health's surgical instruments to enable the development of a powerful AI method that can be used to identify people's behavior and to protect ciphertext data [76].”

• cites [76] Gong T. Huang H. Li P. Zhang K. Jiang H. A medical healthcare system for privacy protec- tion based on iot Proceedings of the 2015 Seventh International Symposium on Parallel Architectures, Algorithms and Programming (PAAP) December 2015 Nanjing, China IEEE 217 222 10.1109/paap.2015.48 2-s2.0-84962254232 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

The authors of [71] propose that unnecessary changes be deleted by examining the importance of small variations in the FL structure for every EHR user, to increase the quality and precision of integration.

“The authors of [71] propose that unnecessary changes be deleted by examining the importance of small variations in the FL structure for every EHR user, to increase the quality and precision of integration.”

• cites [71] Chen H. Li H. Xu G. Zhang Y. Luo X. Achieving privacy-preserving federated learn- ing with irrelevant updates over e-health appli- cations Proceedings of the ICC 2020-2020 IEEE International Conference on Communications (ICC) 2020 Dublin, Ireland IEEE 1 6 10.1109/icc40277.2020.9149385 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Liu et al. built a distributed NN activity plan which enables any clinic to know and understand a part of the model from its EHR source.

“Liu et al. [67] built a distributed NN activity plan which enables any clinic to know and understand a part of the model from its EHR source.”

• cites [67] Liu D. Miller T. Sayeed R. Mandl K. D. Fadl: Federated-autonomous deep learn- ing for distributed electronic health record 2018 arXiv preprint arXiv:1811.11400Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

The role of electronic health records in the healthcare industry is growing rapidly.

“The role of electronic health records in the healthcare industry is growing rapidly.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The integration of autonomous position monitoring and conventional nonconvex enhancement principles builds a different innovative method for FL, aiming to save network resources by interacting with the remote EHR server.

“The integration of autonomous position monitoring and conventional nonconvex enhancement principles [68] builds a different innovative method for FL. The aim is to save network resources by interacting with the remote EHR server.”

• cites [68] Lu S. Zhang Y. Wang Y. Mack C. Learn electronic health Records by Fully De- centralized federated learning 2019 arXiv preprint arXiv:1912.01792Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The next generation of FL solution is cloud edge-based healthcare, which could address issues that directly affect the FL process, such as device, data, and model.

“The next generation of FL solution is presented in [74] cloud edge-based healthcare. This solution could be used to address issues that directly affect the FL process, such as device, data, and model.”

• cites [74] Wu Q. He K. Chen X. Personalized fed- erated learning for intelligent iot applications: a cloud-edge based framework IEEE Open Jour- nal of the Computer Society 2020 1 35 44 10.1109/ojcs.2020.2993259 32396074 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Recent studies reveal that FL is helpful in preventing infections like COVID-19.

“Recent studies reveal that FL is helpful in preventing infections like COVID-19 [78].”

• cites [78] Nguyen D. C. Ding M. Pathirana P. N. Seneviratne A. Blockchain and ai-based solu- tions to combat coronavirus (COVID-19)-like epi- demics: a survey IEEE Access 2021 9 95 730 795 753 10.1109/access.2021.3093633 PMC8545197 34812398 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

FL's utility in the smart healthcare sector with sophisticated features has been proved in recent research.

“FL's utility in the smart healthcare sector with sophisticated features has been proved in recent research.”

May combine two claims.

Pang et al. created a federated model based on a digital city twin concept to predict outcomes of different COVID-19 prevention plans over time and assess the survival rates of multiple cities over the study period.

“Pang et al. [91] created a federated model based on a digital city twin concept to predict outcomes of different COVID-19 prevention plans over time and assess the survival rates of multiple cities over the study period.”

• cites [91] Pang J. Huang Y. Xie Z. Li J. Cai Z. Col- laborative city digital twin for the COVID-19 pan- demic: a federated learning solution Tsinghua Science and Technology 2021 26 5 759 771 10.26599/tst.2021.9010026 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Goldfarb et al. consolidate existing case studies and identify the future challenges in defining FL's benefits and impact on healthcare applications, as well as the obstructions and concerns surrounding its adoption in healthcare.

“Goldfarb et al. [89] consolidate existing case studies and identify the future challenges in defining FL's benefits and impact on healthcare applications, as well as the obstructions and concerns surrounding its adoption in healthcare.”

• cites [89] Goldfarb D. S. Benstein J. A. Zhdanova O. Impending shortages of kidney replacement therapy for COVID-19 pa- tients Clinical Journal of the American Society of Nephrology 2020 15 6 880 882 10.2215/cjn.05180420 32345750 PMC7274293 sourceNot checked yet

  Not checked in the free run.

May combine two claims.

In COVID-19, FL demonstrates its value in diagnosis, treatment, and prognosis prediction.

“In COVID-19, FL demonstrates its value in diagnosis, treatment, and prognosis prediction.”

May combine two claims.

Xu et al. demonstrated that with the help of chest computed tomography (CT) scan data from several institutions in Wuhan, China, they were able to overcome data scarcity, isolation, and heterogeneity and achieve improved detection sensitivity.

“Xu et al. [19] demonstrated that with the help of chest computed tomography (CT) scan data from several institutions in Wuhan, China, they were able to overcome data scarcity, isolation, and heterogeneity and achieve improved detection sensitivity.”

• cites [19] Xu Y. Ma L. Yang F. A col- laborative online ai engine for Ct-based COVID-19 diagnosis medRxiv 2020Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The software and hardware manufacturers would get a lot of benefits from the FL because collaborative learning is possible between hardware devices and software applications without information leakage and the other benefit will come in terms of validation of the AI-based model since it is combined with the hardware devices that helps in continuous improvement of AI-based models.

“The software and hardware manufacturers would get a lot of benefits from the FL because collaborative learning is possible between hardware devices and software applications without information leakage and the other benefit will come in terms of validation of the AI-based model since it is combined with the hardware devices that helps in continuous improvement of AI-based models [110].”

• cites [110] Gadekallu T. R. Pham Q.-V. Nguyen D. C. Blockchain for edge of things: applications, opportunities, and challenges IEEE Internet of Things Journal 2021 9 2 964 988 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

Based on the observations from several states of the art, the FL algorithm analyzes electronic health records to describe and estimate hospitalization, survival, and death rates for people associated with physicians.

“Based on the observations from several states of the art, the FL algorithm analyzes electronic health records to describe and estimate hospitalization, survival, and death rates for people associated with physicians.”

Classified as probably not a substantial claim. May combine two claims.

FL can integrate information acquired and maintained by different institutions in order to capture broader data variability and analyze patients from diverse demographics.

“FL can integrate information acquired and maintained by different institutions in order to capture broader data variability and analyze patients from diverse demographics.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Using the digital city twin platform, they were able to track the effectiveness of each prevention plan and create local models which were sent to the federated sites for safety.

“Furthermore, using the digital city twin platform, they were able to track the effectiveness of each prevention plan and create local models which were sent to the federated sites for safety.”

• cites [91] Pang J. Huang Y. Xie Z. Li J. Cai Z. Col- laborative city digital twin for the COVID-19 pan- demic: a federated learning solution Tsinghua Science and Technology 2021 26 5 759 771 10.26599/tst.2021.9010026 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Medical digital twins can make life-saving decisions on the basis of real-time and historical data with the help of AI.

“Medical digital twins can make life-saving decisions on the basis of real-time and historical data with the help of AI [106, 107].”

• cites [106] El Saddik A. Digital twins: the convergence of multimedia technologies IEEE multimedia 2018 25 2 87 92 10.1109/mmul.2018.023121167 2-s2.0-85051272175 sourceNot checked yet

  Not checked in the free run.

• cites [107] Ross D. Digital twinning [virtual reality avatars] Engineering & Technology 2016 11 4 44 45 10.1049/et.2016.0403 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

This kind of decentralized research could benefit greatly from integrating FL into the study design and data analysis to evaluate the quality of care and outcomes, for example, predicting mortality, complications, hospitalizations, and adverse drug reactions.

“This kind of decentralized research could benefit greatly from integrating FL into the study design and data analysis to evaluate the quality of care and outcomes, for example, predicting mortality, complications, hospitalizations, and adverse drug reactions.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The FL-based technique would be able to help the healthcare providers because it can improve the accuracy and robustness of the AI model and that helps to make the model more generalizable so that it could be used in the real-time environment.

“The FL-based technique would be able to help the healthcare providers because it can improve the accuracy and robustness of the AI model and that helps to make the model more generalizable so that it could be used in the real-time environment [109].”

• cites [109] Aich S. Sinai N. K. Kumar S. Protecting personal healthcare record using blockchain & federated learning technologies Proceedings of the 2022 24th International Conference on Advanced Communication Technology (ICACT) 07-10 February 2021 PyeongChang, Korea (South) IEEE 109 112 sourceNot checked yet

  Not checked in the free run.

Classified as probably not a substantial claim.

Based on their findings, an FL-based architecture would allow global participants to benefit from a globally distributed and real-time CT-COVID-19 diagnostic tool.

“Based on their findings, an FL-based architecture would allow global participants to benefit from a globally distributed and real-time CT-COVID-19 diagnostic tool.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

The suitable use of FL produces generalizable models that will help the achievement of equitable, effective, and patient-centered care.

“The suitable use of FL produces generalizable models that will help the achievement of equitable, effective, and patient-centered care.”

May combine two claims.

FL is primarily designed to safeguard confidentiality by discussing method updates instead of information, although it does not solve many security issues but, like any ML algorithm, always carries some risk.

“FL is primarily designed to safeguard confidentiality by discussing method updates instead of information, although it does not solve many security issues but, like any ML algorithm, always carries some risk.”

May combine two claims.

Given the fact that there can be no comprehensive information distribution among institutions, FL can be trained using non-IID data.

“Given the fact that there can be no comprehensive information distribution among institutions [56, 62], FL can be trained using non-IID data.”

• cites [56] Li W. Milletarì F. Xu D. Privacy-preserving feder- ated brain tumour segmentation Proceedings of the International Workshop on Machine Learning in Medical Imag- Ing 2019 China Springer 133 141 10.1007/978-3-030-32692-0_16 PMC7616593 39453415 sourceNot checked yet

  Not checked in the free run.

• cites [62] Linardos A. Kushibar K. Walsh S. Gkontra P. Lekadir K. Federated learning for multi- center imaging diagnostics: A study in cardiovas- cular disease 2021 arXiv preprint arXiv:2107.03901 PMC8894335 35241683 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

Based on the observations from several states of the art, in an outbreak when facilities may be restricted, it is important to develop diagnostic methods that will help individuals at greater risk of contracting COVID-19 judiciously.

“Based on the observations from several states of the art, in an outbreak when facilities may be restricted, it is important to develop diagnostic methods that will help individuals at greater risk of contracting COVID-19 judiciously.”

May combine two claims.

The privacy and security issues associated with data analysis within a digital twin are addressed by keeping users' data localized within an FL model, which provides the desired outcomes in terms of implementing data analytics.

“The privacy and security issues associated with data analysis within a digital twin are addressed by keeping users' data localized within an FL model, which provides the desired outcomes in terms of implementing data analytics [108].”

• cites [108] Ramu S. P. Boopalan P. Pham Q.-V. Federated learning enabled digital twins for smart cities: concepts, recent Ad- vances, and future directions Sustainable Cities and Society 2022 79 103663 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

FL allows the addition of multiexpert annotation and data from multiple sites acquired with different instruments and techniques.

“Furthermore, FL allows the addition of multiexpert annotation and data from multiple sites acquired with different instruments and techniques.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based). May combine two claims.

FL has been able to harness all of the full learning power of existing data to offer data-driven insights and personalized recommendations due to the uncertainty of long-term complications of COVID-19, the effectiveness of medical treatments, the safety of vaccines, and immunity protection.

“FL has been able to harness all of the full learning power of existing data to offer data-driven insights and personalized recommendations due to the uncertainty of long-term complications of COVID-19, the effectiveness of medical treatments, the safety of vaccines, and immunity protection.”

May combine two claims.

FL can effectively address the issue of data unavailability and get a shared model without obtaining local data.

“FL can effectively address the issue of data unavailability and get a shared model without obtaining local data.”

Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

Using FL in decentralized research would benefit researchers by improving study design and analyzing data on outcomes and quality of care, including predicting mortality, complications, and adverse drug reactions.

“We also find that using FL in decentralized research would benefit researchers by improving study design and analyzing data on outcomes and quality of care, including predicting mortality, complications, and adverse drug reactions.”

Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

FL poses many privacy-preserving challenges in terms of alternatives, techniques, and consequences.

“FL poses many privacy-preserving challenges in terms of alternatives, techniques, and consequences.”

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

FL provides smart healthcare solutions and reshapes current healthcare systems by improving privacy for users and reducing latency among healthcare providers and users alike.

“FL provides smart healthcare solutions and reshapes current healthcare systems by improving privacy for users and reducing latency among healthcare providers and users alike.”

Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based). May combine two claims.

FL does have some advantages, but it also poses challenges that must be considered when setting up federated training efforts.

“FL does have some advantages, but it also poses challenges that must be considered when setting up federated training efforts.”

May combine two claims.

FL has been found to enhance scalability, improve accuracy, reduce training time, improve performance, increase privacy and enhance safety, among others.

“In this context, FL has been found to enhance scalability, improve accuracy, reduce training time, improve performance, increase privacy and enhance safety, among others [105].”

• cites [105] Nguyen D. C. Ding M. Pathirana P. N. Federated Learning for Industrial Internet of Things in Future Industries IEEE Wireless Communications 2021 28 sourceNot checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Boyi Liu et al. proposed an experiment in which four popular federated ML models (Mobile Net, ResNet18, MobileNet-v2, and COVID-Net) were applied to CXR images of patient's chests to compare their performance, and the authors developed these models to detect COVID-19 pneumonia, using the same parameters for all models.

“Boyi Liu et al. [90] proposed an experiment in which four popular federated ML models (Mobile Net, ResNet18, MobileNet-v2, and COVID-Net) were applied to CXR images of patient's chests to compare their performance. The authors developed these models to detect COVID-19 pneumonia, using the same parameters for all models.”

• cites [90] Liu B. Yan B. Zhou Y. Yang Y. Zhang Y. Experiments of federated learning for COVID-19 chest X-ray images 2020 arXiv preprint arXiv:2007.05592Not checked yet

  Not checked in the free run.

May combine two claims.

In non-IID conditions, techniques such as FedAvg frequently fail, which violates the fundamental aim of active learning.

“In such conditions, techniques such as FedAvg [47] frequently fail, which violates the fundamental aim of active learning.”

• cites [47] McMahan B. Moore E. Ramage D. Hamp- son S. y Arcas B. A. Communication-efficient learning of deep networks from decentralized data Artificial Intelligence and Statistics 2017 PMLR 1273 1282Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Researchers and developers get a lot of benefits from the FL-based services because a huge amount of real data would be available to them.

“Researchers and developers get a lot of benefits from the FL-based services because a huge amount of real data would be available to them [111].”

• cites [111] Gadekallu T. R. Pham Q.-V. Huynh-The T. Bhattacharya S. Maddikunta P. K. R. Liyanage M. Federated learning for big data: A survey on opportunities, applications, and future directions 2021 arXiv preprint arXiv:2110.04160Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

Collaborative robotic (Cobot) technology has been widely adopted both by healthcare professionals and by those involved in the medical device industry as a tool to improve workforce efficiency, facilitate workflow improvements, and streamline safety procedures.

“Collaborative robotic (Cobot) technology has been widely adopted both by healthcare professionals and by those involved in the medical device industry as a tool to improve workforce efficiency, facilitate workflow improvements, and streamline safety procedures [112].”

• cites [112] Maddikunta P. K. R. Pham Q.-V. Prabadevi B. Liyanage M. Industry 5.0: a survey on enabling technologies and potential applications Journal of Industrial Information Integration 2021 26Not checked yet

  Not checked in the free run.

Probably not a substantial claim: may not pass S3 (load-bearing).

Implementing FL-based edge intelligence in 6G will bring improved performance.

“However, implementing FL-based Edge intelligence in 6G will bring in improved performance, ultralow latency service, and enhanced privacy of the system [114].”

• cites [114] Nayak S. Patgiri R. 6g communication tech- nology: a vision on intelligent healthcare Health Informatics: A Computational Perspective in Healthcare 2021 Germany Springer 1 18 10.1007/978-981-15-9735-0_1 sourceNot checked yet

  Not checked in the free run.

• AssumptionRests on an assumption: A large number of IoMT devices with massive data in the 6G era will force individuals to deploy efficient machine learning and AI-based algorithms to provide high-quality services.

Probably not a substantial claim: may not pass S2 (checkable), S4 (based).

The federated learning approach is a promising method for achieving strong, precise, safe, robust, and unbiased modeling results.

“FL approach is a promising method for achieving strong, precise, safe, robust, and unbiased modeling results.”

• AssumptionRests on an assumption: Federated learning facilitates communication between various parties without the need to communicate or centralize data, thereby solving difficulties associated with critical health information.

Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

The FL-based Wearable, explainable AI frameworks will enable the user to have better communication using knowledge-based methods and also improve user acceptance and task performance.

“Also, the FL-based Wearable, explainable AI frameworks will enable the user to have better communication using knowledge-based methods and also improve user acceptance and task performance [113].”

• cites [113] Wang S. Qureshi M. A. Miralles-Pechuaán L. Huynh-The T. Gadekallu T. R. Liyan- age M. Explainable ai for b5g/6g: technical as- pects, use cases, and research challenges 2021 arXiv preprint arXiv:2112.04698 sourceNot checked yet

  Not checked in the free run.