CoChat
Sign Up
Cited by 113IoT and Edge/Fog ComputingInternet of Things and AIArtificial Intelligence in Healthcare

The DOI matches its Crossref record; 97 of the 98 references it cites were found in a registry or an index; none of the 107 sources its claims cite is checked for support yet.

Edge AI for Early Detection of Chronic Diseases and the Spread of Infectious Diseases: Opportunities, Challenges, and Future Directions

Open access
Badidi, Elarbi·2023·Future Internet
DOI 10.3390/fi15110370
Open this paper in CoChat — free accountOpen PDF

Abstract

Edge AI, an interdisciplinary technology that enables distributed intelligence with edge devices, is quickly becoming a critical component in early health prediction. Edge AI encompasses data analytics and artificial intelligence (AI) using machine learning, deep learning, and federated learning models deployed and executed at the edge of the network, far from centralized data centers. AI enables the careful analysis of large datasets derived from multiple sources, including electronic health records, wearable devices, and demographic information, making it possible to identify intricate patterns and predict a person’s future health. Federated learning, a novel approach in AI, further enhances this prediction by enabling collaborative training of AI models on distributed edge devices while maintaining privacy. Using edge computing, data can be processed and analyzed locally, reducing latency and enabling instant decision making. This article reviews the role of Edge AI in early health prediction and highlights its potential to improve public health. Topics covered include the use of AI algorithms for early detection of chronic diseases such as diabetes and cancer and the use of edge computing in wearable devices to detect the spread of infectious diseases. In addition to discussing the challenges and limitations of Edge AI in early health prediction, this article emphasizes future research directions to address these concerns and the integration with existing healthcare systems and explore the full potential of these technologies in improving public health.

Overview

AI snapshot

Written from the paper’s full text.

TL;DR

This review examines Edge AI—AI, machine learning, deep learning, and federated learning deployed at the network edge—for early detection of chronic and infectious diseases, arguing it can enable real-time, privacy-preserving health prediction. It surveys applications, models, and challenges, concluding that privacy, data quality, bias, interoperability, and integration with existing health systems must be addressed for adoption.

Key findings

  • Edge AI can process wearable, medical-device, and EHR data locally for real-time early prediction, while federated learning enables collaborative model training without centralizing patient data. One cited example is the EXAM federated learning model, which predicted COVID-19 oxygen requirements using data from 20 institutions and achieved a mean AUC of 0.92 for outcomes within 24 and 72 h.
  • Reported model performances in reviewed studies include random forest achieving 82.26% accuracy for diabetes prediction (Naive Bayes worst at 70.56%); a decision-tree classifier achieving 98% accuracy for heart disease prediction and 99% for classifying subject position; Alzheimer’s prediction reaching 83% test accuracy; and NYUTron achieving an AUC of 78.7–94.9%, improving AUC by 5.36–14.7% over traditional models.
  • Federated learning studies cited include a hybrid classifier-based FL framework improving cardiovascular disease prediction accuracy by 1.5% with 1.6% fewer classification errors; heart disease FL reaching around 89% accuracy on the UCI benchmark dataset; dynamic federated meta-learning improving rare disease prediction accuracy by 13.28%; and COVID-19 mortality FL models outperforming local models (LASSO federated model at three hospitals; MLP federated model at all five hospitals).
  • The paper identifies privacy/security, data quality/accuracy, model bias, interoperability, and integration with existing health systems as key challenges, with future directions including privacy-preserving algorithms, standardization, bias mitigation, and APIs/middleware.

Methods

  • Systematic literature review using keywords and search criteria C1–C4, searching Scopus, Google Scholar, and PubMed, restricted to articles published between 2019 and 2023, and organized around research questions RQ-1 to RQ-4.
  • The initial search found 556 references from Scopus, 565 from Google Scholar, and 198 from PubMed (total 1319); after eliminating duplicates and incomplete references, 280 remained; further title, abstract, and full-text screening eliminated 182, leaving 98 eligible references (excluding background citations).
  • Analysis is a narrative synthesis with summary tables of Edge AI and federated learning applications, methods, results, and limitations.

Limitations

  • Stated: Edge AI for early health prediction faces privacy and security risks, data quality and accuracy problems, model bias, interoperability issues, and challenges integrating with existing health systems.
  • Stated: The review’s methodology did not aim for complete coverage, and the search was restricted to 2019–2023 and three databases. Included federated learning studies often lack details on dataset size/diversity, generalizability beyond specific datasets (e.g., MIMIC-III), real-world examples, and a unified classification of FL challenges.
  • Evident from the text: This is a review, not a primary validation study; reported performance numbers come from cited studies and are not independently reproduced, and no meta-analysis or formal risk-of-bias assessment is described.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3390/fi15110370

Cites

97 of 98 references verified0 citation errors · 1 not checkable

Statistics

Nothing to recompute.

Verification

Claims the paper makes

0 of 107 cited sources checked · 0 supported · 0 flagged

Show all 60 claims

  • Probably not a substantial claim. 59
  • It may state two separate claims as one. 48
  • The quote only partly states this claim. 3
88 of 88

  • The paper relies on this without support.

    Scopus, Google Scholar, and PubMed provide comprehensive coverage of scholarly works on the subject.

    “Three popular databases (Scopus, Google Scholar, and PubMed) renowned for their comprehensive coverage were used to search for scholarly works on the subject.”

  • The paper relies on this without support.

    The search criteria C1-C4 and the 2019-2023 date restriction are adequate to identify relevant literature.

    “The search strategy was based on the above search criteria.A time constraint was applied, restricting the search to include articles published between 2019 and 2023.”

  • The paper relies on this without support.

    Edge devices have limited computational resources compared to cloud-based servers.

    “Edge devices, such as smartphones, IoT devices, or edge servers, often have limited computational resources compared to cloud-based servers.”

  • The paper relies on this without support.

    Patients are the primary users of Edge AI systems for early health prediction and are responsible for collecting and transmitting data to healthcare providers.

    “Patients are the primary users of Edge AI systems for early health prediction and are responsible for collecting and transmitting data to healthcare providers.”

  • The paper relies on this without support.

    Averaging is the most widely used aggregation method in federated learning.

    “Averaging is the most widely used aggregation method in federated learning.”

  • The paper relies on this without support.

    The final 98 references are representative of the literature on Edge AI for early health prediction.

    “The final number of references eligible for this study was 98.”

  • The paper relies on this without support.

    Edge AI relies on collecting vast amounts of personal health data, including medical records, biometric data, and lifestyle information.

    “Edge AI relies on collecting vast amounts of personal health data, including medical records, biometric data, and lifestyle information.”

  • The paper relies on this without support.

    Obtaining informed consent from individuals for data collection and usage can be challenging.

    “Obtaining informed consent from individuals for data collection and usage can be challenging.”

  • The paper relies on this without support.

    Achieving complete anonymity while maintaining data utility for accurate predictions can be a complex task.

    “However, achieving complete anonymity while maintaining data utility for accurate predictions can be a complex task.”

  • The paper relies on this without support.

    Edge AI systems store and process sensitive health data, making them potential targets for cyberattacks.

    “Edge AI systems store and process sensitive health data, making them potential targets for cyberattacks.”

  • The paper relies on this without support.

    Edge AI systems often rely on transmitting data between devices and cloud servers.

    “Edge AI systems often rely on transmitting data between devices and cloud servers.”

  • The paper relies on this without support.

    Edge AI models can be vulnerable to adversarial attacks, where malicious actors manipulate input data to deceive the system.

    “Edge AI models can be vulnerable to adversarial attacks, where malicious actors manipulate input data to deceive the system.”

  • The paper relies on this without support.

    Obtaining ground truth labels for early health prediction can be challenging.

    “However, in early health prediction, obtaining ground truth labels can be challenging.”

  • The paper relies on this without support.

    Ensuring that different devices, systems, and data sources work together is a challenge.

    “Ensuring that different devices, systems, and data sources work together is a challenge.”

  • The paper relies on this without support.

    In early health prediction, data are collected from various sources such as wearable devices, electronic health records, and sensors.

    “In early health prediction, data are collected from various sources such as wearable devices, electronic health records, and sensors.”

  • The paper relies on this without support.

    These sources often use different data formats, protocols, and standards, making it challenging to integrate and analyze data seamlessly.

    “These sources often use different data formats, protocols, and standards, making it challenging to integrate and analyze data seamlessly.”

  • The paper relies on this without support.

    Lack of interoperability can hinder the process of aggregating and fusing data from multiple sources, leading to difficulties in harmonizing and combining heterogeneous data effectively.

    “However, the lack of interoperability can hinder this process, leading to difficulties in harmonizing and combining heterogeneous data effectively.”

  • The paper relies on this without support.

    Edge AI systems rely on efficient communication between edge devices, cloud infrastructure, and central servers.

    “Edge AI systems rely on efficient communication between edge devices, cloud infrastructure, and central servers.”

  • The paper relies on this without support.

    Interoperability issues can arise due to differences in communication protocols, network connectivity, and compatibility between devices and systems.

    “Interoperability issues can arise due to differences in communication protocols, network connectivity, and compatibility between devices and systems.”

  • The paper relies on this without support.

    Interoperability can raise privacy and security concerns, especially when sensitive health data are shared or exchanged between different systems.

    “Interoperability can raise privacy and security concerns, especially when sensitive health data are shared or exchanged between different systems.”

  • The paper relies on this without support.

    Many healthcare organizations have established legacy systems and infrastructure that may not be designed to integrate with newer Edge AI technologies.

    “Many healthcare organizations have established legacy systems and infrastructure that may not be designed to integrate with newer Edge AI technologies.”

  • The paper relies on this without support.

    These legacy systems often have different data formats, protocols, and interfaces, making it difficult to seamlessly incorporate Edge AI solutions.

    “These systems often have different data formats, protocols, and interfaces, making it difficult to seamlessly incorporate Edge AI solutions.”
22 of 22

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?

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.

  • Review table includes an article outside the stated search period

    Minor

    The methodology says the search was restricted to articles published from 2019 to 2023, but Table 1 includes a 2014 article. The table also lists articles from 2015 and 2018.

  • Publication year for reference [85] differs between table and reference list

    Minor

    Table 2 dates reference [85] to 2022, while its reference-list entry identifies the proceedings as 2023.

2 of 2

Citations

Cites

3 retracted

Flagged first, then by citations

Show all 96 references

Research works printed but not linked: 3

  1. Manjulatha, B., and Pabboju, S. (2021). Smart Computing Techniques and Applications, Springer.
  2. Vaid, A., Jaladanki, S.K., Xu, J., Teng, S., Kumar, A., Lee, S., Somani, S., Paranjpe, I., De Freitas, J.K., and Wanyan, T. (2020). Federated Learning of Electronic Health Records Improves Mortality Prediction in Patients Hospitalized with COVID-19. medRxiv, Preprint. · DOI · Paper page
  3. Vaughn, J., Baral, A., Vadari, M., and Boag, W. (2020, January 2–4). Dataset Bias in Diagnostic AI systems: Guidelines for Dataset Collection and Usage. Proceedings of the ACM Conference on Health, Inference and Learning, Toronto, ON, Canada.

Cited by

Sorted by citations

Items 101–113 of 113

Related

See all

Metadata

Authors
Badidi, E.
Affiliations
United Arab Emirates University (AE)
Corresponding author
Badidi, Elarbi
ORCID
Badidi, Elarbi
Venue
Future Internet
Publisher
MDPI AG
Published
2023-11-18
Volume
15
Issue
11
Pages
370
ISSN
1999-5903 (ISSN-L)
DOI
10.3390/fi15110370
OpenAlex
W4388796008
Semantic Scholar
155b078fafb8ee9c25a6274009a315dbb26933ea
Cited by
113
References
96
Citations by year
2026: 37 · 2025: 54 · 2024: 19 · 2023: 3
FWCI
12.54
Type
journal article
Language
en
Open access
Licence
Locations
  • Future Internet (publishedVersion, cc-by)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
United Arab Emirates (UAE) University (G00003443)
Topics
  • Physical Sciences › Computer Science › Computer Networks and Communications › IoT and Edge/Fog Computing
  • Physical Sciences › Computer Science › Information Systems › Internet of Things and AI
  • Health Sciences › Health Professions › Health Information Management › Artificial Intelligence in Healthcare
Keywords
edge AI, edge computing, federated learning, early disease detection, chronic disease, infectious disease surveillance, wearable devices, diabetes mellitus, cancer, electronic health records, privacy-preserving machine learning, healthcare integration

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.

A language model drew the claims from the paper on 2026-10-09, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-09 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

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