About the Journal

About the ASIDE Artificial Intelligence in Healthcare

Vision and Mission

ASIDE Artificial Intelligence in Healthcare is a peer-reviewed, open-access journal committed to advancing diversity, equity, inclusion, and accessibility in health and healthcare through scholarly publication. As part of ASIDE Healthcare, the journal reflects these values in all activities and publications.

Our mission is to advance health artificial intelligence by promoting innovative research, evidence-informed clinical practice, and transformative ideas that prioritize health equity. We serve as a platform for voices historically underrepresented in medical research and healthcare delivery, supporting scholarship that improves health outcomes across diverse communities.

Scope

ASIDE Artificial Intelligence in Healthcare is a peer-reviewed, open-access journal publishing research on the development, evaluation, deployment and governance of artificial intelligence and machine learning across health and healthcare. The journal welcomes work across the full translational path, from model development to real-world clinical impact.

The journal considers manuscripts in the following areas:

  • Clinical AI and decision support — diagnostic, prognostic and treatment-recommendation models; clinical deployment and workflow integration; human–AI interaction and clinician trust.
  • Large language models and generative AI in health — clinical documentation, patient communication, triage, summarisation, coding, and evaluation of factuality and safety.
  • Medical imaging and signal analysis — radiology, pathology, dermatology, ophthalmology, cardiology and physiological signal interpretation.
  • Predictive analytics and health-system operations — risk stratification, capacity and resource planning, quality and safety monitoring, population health.
  • Clinical informatics and data infrastructure — electronic health record data quality, interoperability, federated learning, privacy-preserving methods, real-world data and evidence.
  • Digital health and remote monitoring — wearables, sensors, telehealth and mobile health where AI or advanced analytics is central.
  • Drug discovery, genomics and precision medicine — applications with a clear clinical or translational link.
  • Evaluation science and methodology — prospective and randomised evaluations, external validation, calibration and drift, silent trials, health-economic evaluation, implementation science.
  • Fairness, equity and bias — algorithmic bias detection and mitigation, performance across demographic subgroups, representation in training data, impact on underserved populations.
  • Ethics, regulation and governance — transparency, accountability, consent, liability, regulatory pathways, post-market surveillance and institutional AI governance.
  • Education and workforce — AI literacy and curricula for clinicians, trainees and patients.

Consistent with the mission of the American Society for Inclusion, Diversity, and Equity in Healthcare, the journal gives particular priority to studies that examine equity, bias and performance in underserved, minoritised and resource-limited populations, and to work from low- and middle-income settings. Negative results, failed deployments, external validations that do not replicate original performance, and replication studies are explicitly welcomed.

Out of scope: purely technical machine-learning methodology with no health application; commercial product descriptions without independent evaluation; and studies reporting only internal-split performance of a model with no external validation, clinical context or stated intended use.

Publication Frequency

ASIDE Artificial Intelligence in Healthcare is published quarterly (four issues per year). Each issue contains peer-reviewed original research articles, reviews, case reports, and other scholarly content in artificial intelligence and health.

Readership

Our readers include clinicians, informaticians, data scientists, methodologists, researchers, allied health professionals, educators, regulators and policy makers dedicated to the safe and equitable use of artificial intelligence in health for all. With a global reach, the journal invites submissions worldwide to foster a diverse and clinically meaningful dialogue in health artificial intelligence.

Open Access & Licensing Policy

ASIDE Artificial Intelligence in Healthcare provides immediate open access to all content. The full text of every article is freely available to read, download, copy, distribute, print, search, or link to without delay, with no embargo and no registration requirement.

Unless otherwise indicated, all articles are published under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. This license permits use, sharing, adaptation, distribution, and reproduction in any medium or format, including commercial use, provided appropriate credit is given to the original authors and the source, a link to the license is provided, and changes are indicated.

Authors retain copyright to their work and grant the publisher a non-exclusive right of first publication under the above license. The copyright holder and license are clearly displayed on each article’s HTML page and PDF.

Content (e.g., images, figures, data) not covered by CC BY 4.0 is identified by a credit line noting the different terms. Permission may be required for uses beyond those allowed by the stated license.

The journal charges no submission or publication fees (no APCs).

Editorial Process

We uphold a rigorous peer-review process managed by an international editorial board committed to advancing health artificial intelligence and upholding fairness, diversity, and inclusiveness in academic publishing.

ASIDE Artificial Intelligence in Healthcare maintains editorial independence by not accepting any form of advertising.

Direct Marketing Policy

ASIDE Artificial Intelligence in Healthcare may occasionally send targeted calls for papers or informational messages to researchers whose expertise aligns with the journal’s aims and scope. Any outreach conducted on behalf of the journal is appropriate, well-targeted, and unobtrusive. All information about the journal and publisher is truthful and not misleading.

We do not guarantee acceptance, indexing, or publication outcomes in any communication, and we do not request or require payment as a condition of submission. Messages clearly identify the journal and publisher, the article types sought, relevant deadlines, and link to our policies. Recipients can opt out at any time, and we promptly honor unsubscribe requests. We do not buy third-party email lists; contact details are gathered from publicly available academic sources or prior voluntary interactions. Outreach practices comply with applicable regulations (e.g., CAN-SPAM/GDPR).

Peer Review Policy (Double-Anonymous)

1. Introduction to Peer Review

1.1 What is Peer Review? Independent experts assess submissions for originality, validity, and significance to inform editorial decisions.

1.2 Purpose at ASIDE Artificial Intelligence in Healthcare: To validate scientific accuracy and relevance so that published research is trustworthy and contributes meaningfully to health artificial intelligence.

2. Types of Peer Review

2.1 Double-Anonymous Peer Review Both reviewers and authors remain anonymous to minimize bias related to identities.

3. Double-Anonymous Process

3.1 Manuscript Submission

  • Authors submit via the journal’s system, ensuring manuscripts and supplementary files contain no identifying information.
  • Editorial staff assess scope fit and guideline adherence before review.

3.2 Reviewer Selection

  • Editors invite expert reviewers based on subject expertise, prior review quality, and impartiality.

3.3 Conducting the Review

  • Reviewers evaluate significance, originality, methodology, data analysis, and presentation using predefined criteria.
  • Confidentiality is mandatory; conflicts of interest must be declared.

3.4 Reviewer Recommendations

  • Possible recommendations: accept, minor revisions, major revisions, or reject, with constructive feedback to strengthen the work.

3.5 Decision Making

  • Editors consider all feedback; in case of conflicting opinions, additional reviews may be solicited.

3.6 Revision and Re-review

  • Authors address comments and resubmit; further rounds may occur to ensure concerns are fully resolved.

3.7 Acceptance and Publication

  • Accepted manuscripts proceed to copyediting, layout, and proofing prior to publication.

4. Why Double-Anonymous?

To enhance impartiality and focus evaluation on academic content and scientific merit.

5. Ensuring Integrity and Fairness

Attempts to circumvent anonymity (e.g., identifying authors or reviewers) should be reported and will be addressed seriously.

6. Conclusion

The peer-review process is designed to uphold the highest standards of scientific integrity and publication quality, contributing valuable knowledge to health artificial intelligence.

Artificial Intelligence Policy

A. Reporting standards for research about AI

All submissions reporting an artificial intelligence or machine-learning model must upload the applicable reporting checklist as a supplementary file:

  • Prediction model development or validation — TRIPOD+AI
  • Randomised trial of an AI intervention — CONSORT-AI
  • Trial protocol — SPIRIT-AI
  • Diagnostic accuracy study — STARD-AI
  • Medical imaging AI — CLAIM
  • Early-stage or decision-support evaluation — DECIDE-AI
  • Systematic review — PRISMA 2020
  • Quality improvement or implementation — SQUIRE 2.0 or StaRI

Every manuscript describing a model must state, in the Methods:

  1. Intended use — the clinical task, target population, care setting, and the decision the model is meant to inform.
  2. Data provenance — data sources, collection period, geography, institution type, inclusion and exclusion criteria, and handling of missing data.
  3. Population composition — the demographic composition of training, validation and test sets, including age, sex, race or ethnicity, insurance or socioeconomic indicators, and language where available. Authors who cannot report these must say so and explain why.
  4. Model specification — architecture or algorithm family, inputs and preprocessing, outputs, hyperparameters and version. For commercial or closed models, the exact product name, version and access date.
  5. Evaluation design — how training, validation and test data were separated, with explicit confirmation that there is no patient-level or institution-level leakage between them.
  6. Performance reporting — discrimination and calibration with confidence intervals, plus clinically meaningful measures such as net benefit, number needed to alert or false-alert rate at the operating threshold. The chosen decision threshold and its justification must be stated.
  7. Subgroup performance — performance disaggregated by the demographic groups above. A manuscript reporting only aggregate performance will be returned for revision before review.
  8. Failure modes and limitations — where the model underperforms, known distribution shift or drift, and populations to which results should not be extrapolated.
  9. Comparator — performance against current standard of care, an existing validated model, or a clinician baseline, wherever feasible.

Code and data availability. Analysis code must be deposited in a public repository with a persistent identifier, or authors must state a specific reason why not. Where data cannot be shared for privacy or governance reasons, authors must provide a data dictionary, synthetic or summary data, and clear instructions for legitimate access requests. “Data available on reasonable request” alone is not accepted.

Prospective and external validation. Models evaluated only on retrospective, single-institution, internally split data may be published, but the abstract and conclusions must describe the work as preliminary and must not claim clinical readiness or benefit.

Conflicts of interest. Any commercial relationship with a vendor of a model or platform evaluated in the manuscript must be disclosed, and at least one author with no such relationship should have full access to the data and control over the decision to publish.

B. Use of generative AI by authors

  1. AI cannot be an author. Authorship requires accountability, which a tool cannot hold. AI tools must not be listed as authors or co-authors.
  2. Disclosure is mandatory. Any use of generative AI in the writing, editing, translation, coding, data analysis or figure generation of the manuscript must be declared in the Large-Language Model (LLM) – Generative AI Use Statement, naming the tool, version, date and specific purpose.
  3. Routine assistance is exempt — spelling and grammar checking, reference formatting, and standard statistical software.
  4. Authors are fully responsible for all content, including the accuracy of facts, citations and quotations. Fabricated or inaccurate references generated by AI tools are treated as a serious error and may result in rejection or, after publication, correction or retraction.
  5. AI-generated images and figures representing data, patients or clinical findings are not permitted. Conceptual or explanatory illustrations produced with AI must be labelled as such in the caption.
  6. Patient data must never be entered into public or non-compliant AI tools. Authors must confirm that no identifiable patient information was submitted to any third-party system without appropriate governance approval.

C. Use of AI in peer review and editorial decisions

  1. Reviewers and editors must not upload any part of a confidential manuscript into a public generative AI tool. Doing so breaches confidentiality.
  2. AI tools may be used by reviewers only for language assistance on their own written comments, and this must be disclosed to the editor.
  3. Editorial decisions are made by human editors. AI is used by the editorial office only for administrative screening — plagiarism detection, reference checking, image-integrity screening, statistical and reporting-checklist completeness, and reviewer identification — and never to accept or reject a manuscript.
  4. Where an automated screening tool flags a manuscript, the flag is verified by a human editor before any action is communicated to authors.

Policies on Editorial Involvement and Endogeny

Policy on Editorial Involvement in Manuscripts

Conflict of Interest for Editorial Members

  • If the Editor-in-Chief, an Associate Editor, or an editorial board member is a co-author of a submitted manuscript, they must recuse themselves from handling that manuscript.
  • Another qualified editor with no authorship role will manage the review to ensure unbiased evaluation and decision-making.

Decision Making and Access

  • Editorial members who are co-authors will not have access to the review process for their own manuscripts and will not influence the decision.

Endogeny

The journal seeks to minimize endogeny. The proportion of published research papers in which at least one author is an editor, editorial board member, or reviewer of the journal should not exceed 25% in either of the latest two issues, or the equivalent recent publication period if the journal’s publication model changes.

Ethical Standards

Introduction

ASIDE Artificial Intelligence in Healthcare adheres to the codes and best practices of COPE, ICMJE, OASPA, and the Think. Check. Submit. initiative to ensure integrity, transparency, and fairness. This section clarifies expected ethical behavior for authors, editors, reviewers, and the publisher.

1. Editorial Standards

  • Integrity: We follow COPE and ICMJE guidance to maintain the scholarly record.
  • Fair Play: Manuscripts are evaluated for intellectual content without regard to authors’ personal attributes or beliefs.
  • Confidentiality: Editorial information is disclosed only to parties directly involved in the editorial process.

2. Duties of Reviewers

  • Contribution: Reviews assist editorial decisions and help authors improve their work.
  • Promptness: Reviewers who are unqualified or unavailable promptly notify the editor.
  • Confidentiality: Manuscripts under review are confidential.
  • Objectivity: Reviews should be evidence-based and free of personal criticism.
  • Conflicts of Interest: Reviewers must decline when conflicts exist.

3. Duties of Authors

  • Reporting Standards: Present accurate work and objective discussion of significance.
  • Data Access & Retention: Provide data for editorial review and public access when requested.
  • Originality & Citation: Submit original work with proper citation of sources.
  • Redundant/Concurrent Publication: Do not publish substantially similar work elsewhere.
  • Authorship: Limit authorship to those with significant contributions.
  • Disclosures: Declare financial or substantive conflicts that could influence results or interpretation.

4. Publishing Ethics Issues

  • Monitoring Ethics: The editorial board follows COPE in addressing ethics concerns and correcting the record.
  • Retractions: Confirmed misconduct will result in retraction when warranted.

Authors are encouraged to consult the principles of the Think. Check. Submit. initiative when selecting journals. For questions, contact the editorial office at contact@asidejournals.com.

Preservation & Archiving Policy

ASIDE Artificial Intelligence in Healthcare is committed to the long-term preservation and accessibility of its content through the following:

  • Internet Archive: The journal aims to deposit issues and article PDFs in the Internet Archive (archive.org) for long-term preservation and public access. Core journal pages may also be captured periodically via the Wayback Machine.
  • Institutional & Library Repositories (Self-archiving): Authors are encouraged to deposit all versions—preprint, accepted manuscript, and version of record—without embargo, in institutional or subject repositories of their choice.
  • OAI-PMH: Metadata and content are available for harvesting via our OAI-PMH endpoint: https://asidejournals.com/index.php/artificial-intelligence/oai

If the journal were to cease publication, archived content will remain accessible via the Internet Archive and any active preservation services (e.g., PKP PN), where applicable.

Repository & Self-Archiving Policy

ASIDE Artificial Intelligence in Healthcare permits and encourages authors to deposit all versions of their articles in institutional, subject, funder, or personal repositories, and on academic profiles, without embargo:

  • Submitted version (preprint): may be shared at any time.
  • Accepted version (Author Accepted Manuscript, AAM): may be shared immediately upon acceptance.
  • Published version (Version of Record, VoR): may be shared immediately after publication.

Include a full citation and a link to the Version of Record (article page or DOI) wherever a version is deposited. Articles are published under CC BY 4.0. Journal metadata is harvestable via OAI-PMH: https://asidejournals.com/index.php/artificial-intelligence/oai.

Data Sharing & Reproducibility

ASIDE Artificial Intelligence in Healthcare supports open, reproducible research. All research articles must include a Data Availability Statement describing where the data, code, and materials that support the findings can be accessed, with persistent identifiers (e.g., DOIs) wherever possible. For clinical trials, authors must provide an ICMJE-compliant data sharing statement.

Underlying data (raw and processed), analysis scripts/code, protocols, and other materials necessary to reproduce the results should be shared unless restricted for ethical, legal, or proprietary reasons. Where restrictions apply, authors must explain them and provide conditions for controlled access.

Deposit datasets in a trusted repository—preferably a discipline-specific repository when appropriate. When no domain repository fits, use a generalist repository that issues DOIs (e.g., Zenodo, Dryad, OSF). Authors can locate suitable repositories via the re3data registry. Software/code should be in a public version-controlled repository (e.g., GitHub) and archived with a DOI (e.g., via GitHub→Zenodo integration) for citation.

Use open, non-proprietary formats where feasible (e.g., CSV/TSV, JSON, TXT, PNG/TIFF). Provide sufficient documentation/metadata to make data FAIR—Findable, Accessible, Interoperable, and Reusable (e.g., README files, variable dictionaries, licenses).

Human-participant data must be de-identified and shared in line with informed consent, IRB/ethics approvals, and applicable regulations. If data cannot be made public, deposit metadata and specify a controlled access mechanism or qualified point of contact in the Data Availability Statement.

Data and code should be deposited upon acceptance and publicly available upon publication. Include dataset and software citations (with DOIs) in the reference list and provide repository links in the Data Availability Statement.

Publisher Information