Search
This guide provides essential information for authors preparing manuscripts for submission to the journal. Authors are encouraged to read this guide carefully before submission and to consult the journal’s Aims and Scope, Article Types, Open Access and Fees, and Journal Policies pages where applicable.
The journal publishes scholarly work in medical artificial intelligence, machine learning, deep learning, clinical decision support, diagnostic AI, predictive analytics, medical imaging AI, digital health, natural language processing in medicine, AI-enabled clinical workflows, trustworthy AI, real-world validation, translational medical AI and related clinical, technical, ethical and methodological research.
The journal welcomes studies that advance the development, validation, evaluation, implementation, governance or responsible use of AI in medicine and healthcare. Manuscripts should report AI systems, datasets, methods, validation procedures, limitations, clinical relevance, ethical safeguards and transparency materials with sufficient clarity to allow critical appraisal and, where appropriate, reproducibility.
Manuscripts that do not meet the journal’s requirements may be returned for correction before editorial assessment or peer review. Submission of a manuscript does not guarantee external peer review or acceptance.
Before submitting a manuscript, authors should ensure that the work is suitable for the journal and that all required materials are prepared.
Authors should:
read the journal’s Aims and Scope to confirm that the manuscript fits the journal’s academic focus and target readership;
select the appropriate article type and follow the relevant requirements for structure, word limits, references, tables, figures and supplementary materials;
ensure that the manuscript is original, has not been formally published previously and is not under consideration by another journal;
obtain approval from all authors for the submitted version;
ensure that authorship, author order and corresponding author details have been agreed by all authors before submission;
prepare all required statements, including ethics approval, informed consent, author contributions, competing interests, funding, data availability, clinical trial registration and AI-use statements where applicable;
clearly identify the AI system, model, algorithm, software, dataset, benchmark, clinical decision support tool, digital health technology or medical device studied, where applicable;
disclose data sources, model development procedures, validation approach, intended use, target population, clinical setting, performance metrics, limitations and potential risks where applicable;
disclose sponsor, funder, technology provider, data provider, cloud service provider, software developer or commercial partner involvement in data access, model development, analysis, interpretation, manuscript preparation and publication decisions;
follow appropriate reporting guidelines according to the study design;
obtain permission for any third-party material, including figures, tables, images, screenshots, datasets, code, software interfaces, model outputs, benchmark materials or substantial text reproduced or adapted from other sources;
review the journal’s policies on peer review, publication ethics, authorship, research ethics, data availability, use of AI, copyright, licensing and post-publication updates.
Authors who are unsure whether their manuscript is suitable for the journal may contact the Editorial Office with a pre-submission inquiry. A pre-submission inquiry should include the manuscript title, author list, abstract and a brief explanation of the manuscript’s relevance to the journal. A positive response to a pre-submission inquiry does not guarantee peer review or acceptance.
Preprints
The journal considers manuscripts that have been posted on recognized preprint servers, provided that the work has not been formally published and is not under consideration by another journal at the same time.
Authors must disclose any preprint at submission and provide the preprint server name, DOI or link where applicable. If the manuscript is accepted, authors are encouraged to update the preprint record with a citation and link to the final published article.
Online Submission
Manuscripts must be submitted through the journal’s online submission system. Authors should create a user account, log in and submit the manuscript following the instructions in the system. Submissions sent by email are generally not considered unless specifically requested by the Editorial Office.
The maximum file size for individual uploaded files is 50 MB, unless otherwise specified by the submission system.
The journal considers several types of manuscripts. Authors should select the article type that best matches the purpose, design and scope of their work.
Detailed requirements for each article type, including word limits, abstract format, reference limits, figure and table limits and required sections, are provided on the Article Types page.
Common article types may include:
Original Research Article
Reports original research findings based on clinical, technical, computational, translational, implementation, methodological, epidemiological or other scholarly research in medical AI. Original Research Articles should generally include Introduction, Methods, Results and Discussion sections.
Medical AI Model Development and Validation Study
Reports the development, internal validation, external validation, updating or evaluation of AI models, machine learning models, deep learning systems, prediction models, diagnostic models, prognostic models, risk stratification tools or clinical decision support algorithms. Authors should describe the intended use, data sources, model architecture, training and validation strategy, performance metrics, limitations and clinical relevance.
Diagnostic AI / Medical Imaging AI Study
Reports studies involving AI-enabled diagnosis, image analysis, radiology, pathology, endoscopy, ultrasound, ophthalmology, dermatology, electrocardiography or other signal or image-based medical AI applications. Authors should describe image acquisition, annotation, reference standards, dataset splits, external validation, human comparator, performance metrics and clinical workflow where applicable.
Clinical Trial / AI-enabled Intervention Study
Reports prospective interventional studies involving AI-enabled clinical decision support, diagnostic AI, digital therapeutics, automated triage, remote monitoring, workflow optimization or other AI-enabled healthcare interventions. Authors should provide ethics approval, trial registration information, intervention details, safety monitoring and data sharing information where applicable.
Clinical Decision Support / Digital Health AI Study
Reports studies of AI-enabled tools embedded in clinical workflow, electronic health records, telemedicine platforms, remote monitoring systems, mobile health applications or hospital information systems. Authors should describe intended users, clinical workflow, human oversight, alert logic, implementation context, safety safeguards and evaluation methods.
Dataset, Benchmark or Data Resource Article
Describes a dataset, benchmark, annotation resource, clinical data resource, imaging dataset, synthetic dataset, benchmark task or evaluation platform of clear value to the field. Authors should provide information on data provenance, consent or governance basis, inclusion criteria, data quality, annotation process, access conditions, license, limitations and intended use.
Software, Model or Tool Article
Describes software, model implementations, model cards, decision support tools, libraries, pipelines, validation tools or other reusable technical resources. Authors should provide access information, documentation, license, dependencies, version information, testing or validation evidence and limitations.
Real-world Evaluation / Implementation Study
Reports the evaluation, deployment, monitoring or implementation of AI systems in real-world clinical, operational or public health settings. Authors should describe setting, users, workflow integration, data shift, monitoring, human oversight, adoption, safety, equity, unintended consequences and sustainability.
Review Article
Provides a comprehensive, balanced and critical overview of a defined topic. Review Articles should synthesize current evidence, identify knowledge gaps and discuss future directions.
Systematic Review and Meta-analysis
Presents a systematic search, selection, appraisal and synthesis of existing evidence. Authors should follow appropriate reporting guidelines and provide protocol registration information where applicable.
Method / Technical Note
Describes a novel method, technical improvement, validation approach, evaluation framework, data processing method, model auditing method, fairness assessment, interpretability approach or reproducibility tool relevant to medical AI.
Protocol
Describes a planned clinical trial, model validation study, systematic review, implementation study, dataset study or other research protocol. Authors should provide registration information where applicable and clearly describe the rationale, design, methods and planned analysis.
Brief Report / Short Communication
Reports concise but significant findings that may not require a full-length article. These submissions should present clear methods, focused results and a well-supported conclusion.
Case Report / Case Series
Describes clinically or technically important cases involving AI-enabled clinical decision-making, diagnostic support, digital health tools, algorithmic error, patient safety issue or implementation challenge. Written informed consent for publication must be obtained when identifiable patient information, clinical images, screenshots, case details or outputs are included.
Guideline / Consensus
Provides evidence-based or expert consensus recommendations on a specific topic. The methodology for evidence review, panel composition, conflict-of-interest management, consensus development and recommendation grading should be clearly described.
Editorial / Commentary / Perspective
Presents expert opinion, critical discussion or interpretation of current issues in medical AI, digital health, clinical implementation, AI governance, regulation, ethics or policy. These article types are usually commissioned, but unsolicited submissions may be considered at the editors’ discretion.
Letter to the Editor
Provides brief comments on articles published in the journal or concise observations of interest to readers.
Manuscripts should be prepared clearly, concisely and consistently. Authors should write for an international scholarly audience and avoid unnecessary jargon, unsupported claims or excessive technical detail that is not essential to understanding the work.
File Format
Manuscripts should be submitted in an editable format, such as Microsoft Word. LaTeX files may be accepted where appropriate. Figures, tables, model diagrams, screenshots, code documentation and supplementary materials should be prepared according to the journal’s file and formatting requirements.
Title Page
The title page should include:
manuscript title;
full names of all authors;
author affiliations;
corresponding author name, email address and full contact details;
ORCID iDs, where available;
equal contribution or co-corresponding author notes, where applicable;
short running title, if required;
word count, number of figures, number of tables and number of supplementary files, if requested.
Abstract and Keywords
The abstract should accurately summarize the purpose, methods, main findings and conclusions of the manuscript. Original Research Articles, model development studies, validation studies, clinical studies and implementation studies should generally use a structured abstract where required by the article type.
Authors should provide keywords that are specific, searchable and relevant to the manuscript. Standardized medical, technical and AI-related terms should be used where appropriate.
Main Text
Original Research Articles should generally include:
Introduction;
Methods;
Results;
Discussion;
Conclusions, where appropriate.
The Methods section should provide sufficient detail to allow the work to be evaluated and, where applicable, reproduced. Authors should describe study design, setting, participants or data sources, eligibility criteria, data collection, data preprocessing, model development, validation, comparator, outcome measures, statistical methods, software, sample size considerations, evaluation metrics and analysis plans where relevant.
The Results section should present findings in a logical order, supported by appropriate tables, figures and statistical information. The Discussion should interpret the findings in the context of existing evidence, acknowledge limitations and avoid overstating clinical readiness, safety, effectiveness or generalizability.
AI System, Model and Dataset Details
Where applicable, manuscripts should clearly identify the AI system, model, algorithm, software, dataset, benchmark, platform or clinical decision support tool studied.
Authors should provide sufficient information on:
intended use and clinical context;
target population and intended users;
model name, architecture, version and major components;
software, package, framework or platform version;
data sources, setting, time period and inclusion criteria;
data collection, preprocessing, labeling and quality control;
annotation process, annotator expertise and inter-annotator agreement where relevant;
training, validation and test sets;
measures taken to prevent data leakage;
baseline models and comparator selection;
evaluation metrics and statistical uncertainty;
calibration, discrimination, clinical utility and error analysis where relevant;
subgroup analyses, fairness assessment and bias evaluation where applicable;
external validation and temporal validation where available;
human oversight, workflow integration and safety safeguards;
known limitations, failure modes and conditions under which the system should not be used.
Authors should avoid promotional language and should not present experimental, limited-validation or retrospective results as established clinical effectiveness.
Tables, Figures and Visual Materials
Tables and figures should be numbered consecutively in the order in which they are cited in the text. Each table and figure should have a clear title and legend.
Figures, screenshots, model diagrams and visual materials should be of sufficient resolution for peer review and publication. Authors should avoid image manipulation that could misrepresent the data, model output or clinical interpretation. Adjustments to brightness, contrast or color balance must not obscure, eliminate or misrepresent information.
Where figures, tables, screenshots, software interfaces, model outputs, benchmark materials or other content are reproduced or adapted from another source, authors must obtain appropriate permission and provide proper attribution.
Clinical Images, Screenshots and Model Outputs
Clinical images, radiological images, pathology images, endoscopic images, dermatological images, physiological signals, screenshots, model outputs, dashboards, clinical decision support interfaces and other visual materials should be anonymized before submission where applicable.
Written consent for publication is required when any patient, participant, case detail, image, screenshot, interface, voice, video, interaction log or other material may be identifiable.
Supplementary Materials
Supplementary materials should be directly relevant to the manuscript and should support, but not replace, the main text. Supplementary files may include extended methods, additional tables, figures, datasets, checklists, protocols, statistical analysis plans, model cards, data cards, source code, prompts, configuration files, hyperparameters, annotation guidelines, evaluation scripts or other supporting information.
References
References should be accurate, complete and relevant. Authors are responsible for checking all references before submission. References should be formatted in accordance with the journal’s reference style. Please click here to download the journal’s EndNote output style.
Authors should avoid inappropriate citation practices, including excessive self-citation, citation manipulation or citation of unreliable sources. Authors should verify all references, including those suggested or generated with the assistance of AI tools.
Language and Style
Manuscripts should be written in clear English. Authors should use precise, professional and non-promotional language. Claims regarding model performance, safety, fairness, clinical effectiveness, workflow impact or deployment readiness should be supported by appropriate evidence.
Authors who are not fluent in English may consider professional language editing before submission. Language editing does not guarantee editorial review or acceptance.
Authors should follow appropriate reporting guidelines according to study design and should submit completed checklists where required.
Relevant guidelines may include:
CONSORT for randomized trials;
CONSORT-AI for clinical trials involving AI components;
SPIRIT and SPIRIT-AI for clinical trial protocols involving AI components;
PRISMA for systematic reviews and meta-analyses;
STROBE for observational studies;
STARD and STARD-AI for diagnostic accuracy studies where applicable;
TRIPOD+AI for prediction model development, validation, updating or evaluation;
CLAIM or other AI imaging checklists where appropriate;
CARE for case reports;
ARRIVE for animal studies;
TIDieR for intervention description;
SQUIRE for quality improvement studies.
Authors should consult the EQUATOR Network and relevant field-specific reporting guidance to identify the most appropriate reporting guideline for their study design.
Model Development and Validation
For AI model development and validation studies, authors should report the study design, data source, participant or sample selection, outcome definition, predictors or input features, missing data handling, preprocessing, model development methods, validation strategy, evaluation metrics and limitations.
Authors should clearly distinguish between development, internal validation, external validation, prospective validation, clinical evaluation and real-world implementation. Claims about clinical utility or safety should be supported by appropriate evidence.
Diagnostic, Prognostic and Prediction Model Studies
For diagnostic, prognostic or prediction model studies, authors should describe the target condition or outcome, reference standard, time horizon, intended clinical role, candidate predictors, model development process, calibration, discrimination, clinical utility, decision thresholds, subgroup performance and external validation where available.
Authors should report uncertainty around performance estimates and should avoid interpreting retrospective performance as evidence of clinical effectiveness without appropriate validation.
Medical Imaging and Signal-based AI
For imaging, pathology, endoscopy, dermatology, ECG, EEG, waveform or other signal-based AI studies, authors should describe data acquisition, device or scanner details, image or signal quality control, annotation process, reference standard, assessor expertise, blinding, dataset splits, external validation, performance metrics and clinical relevance.
Authors should clearly report whether images or signals from the same patient, examination, institution or time period could appear in more than one dataset split and how leakage was prevented.
Datasets, Benchmarks and Data Resources
For dataset or benchmark articles, authors should describe data provenance, inclusion criteria, consent or governance basis, sampling strategy, annotation process, data quality, representativeness, missingness, licensing, access conditions, known limitations and intended use.
Authors should discuss potential dataset bias, demographic imbalance, site-specific effects, label noise, data drift, benchmark contamination, overfitting risks and appropriate use restrictions.
Real-world Evaluation and Implementation
For real-world evaluation or implementation studies, authors should describe the clinical setting, workflow, users, deployment environment, implementation strategy, human oversight, monitoring, adoption, alert burden, user behavior, unintended consequences, equity considerations and sustainability.
Authors should report safety outcomes, workflow impact, failure modes, override rates, clinician or patient acceptance and limitations where applicable.
Clinical Trial Registration and Data Sharing
Clinical trials and interventional studies involving AI-enabled diagnostic systems, clinical decision support, digital therapeutics, automated triage, remote monitoring, workflow tools or other AI-enabled healthcare interventions should be registered in a publicly accessible registry before participant recruitment begins, where applicable. The registry name and registration number should be provided in the manuscript.
Manuscripts reporting clinical trials should include a data sharing statement where required by applicable standards or journal policy. The statement should indicate whether de-identified individual participant data will be shared, what data will be shared, whether related documents such as the study protocol or statistical analysis plan will be available, when and for how long data will be available, and the conditions for access.
Responsible AI, Bias, Fairness and Safety
Where applicable, authors should discuss bias, fairness, privacy, robustness, explainability, human oversight, accountability, data shift, automation bias, alert fatigue, cybersecurity, environmental impact, model updating, monitoring and foreseeable misuse.
Authors should avoid overstating AI system capabilities and should distinguish clearly between experimental performance, benchmark performance, clinical validation and real-world deployment readiness.
Authors must include all statements required by the journal and applicable policies. These statements should usually appear in the manuscript file, title page, acknowledgements section or submission system, according to the journal’s instructions.
Required statements may include:
Ethics Approval
Studies involving human participants, human data, clinical data, medical images, biosignals, electronic health records, registry data, administrative data, social media data, behavioral data, voice data, video data, biometric data, location data, animals or identifiable information must include the name of the ethics committee, institutional review board or relevant authority, approval number or reference, and any relevant exemption information.
Informed Consent
For studies involving human participants, authors should state whether informed consent was obtained. For identifiable patient information, case details, clinical images, screenshots, model outputs, voice recordings, videos, interaction logs or any other identifiable material, written consent for publication must be obtained where applicable.
Data Governance and Privacy
For studies using clinical datasets, routinely collected data, electronic health records, registry data, imaging data, public datasets, scraped data or third-party datasets, authors should describe the ethical, legal or governance basis for data use, including consent, waiver of consent, data use agreement, public availability or institutional approval where applicable.
Authors should describe how privacy, confidentiality, de-identification, anonymization, data minimization, access control and participant protection were addressed where applicable.
Clinical Trial Registration
Clinical trials and interventional studies should provide the registry name and registration number where applicable. Registration should occur before participant recruitment begins.
Author Contributions
Authors should describe each author’s contribution to the work. The journal encourages use of the CRediT taxonomy where appropriate.
Competing Interests
All authors must disclose any financial or non-financial competing interests. For medical AI research, relevant disclosures may include employment, consultancy, advisory roles, honoraria, stock ownership, patents, patent applications, licensing agreements, sponsored research, software ownership, model ownership, platform access, cloud credits, compute grants, dataset access, model access, technical support, statistical support, writing assistance or relationships with technology companies, data providers, cloud service providers, platform operators, healthcare organizations, insurers, government agencies or sponsors.
Funding
Authors should disclose all funding sources and grant numbers. Compute grants, cloud credits, data access support, software access, model access and in-kind support should also be disclosed. The role of funders, sponsors, technology providers or data providers in study design, data access, model development, experimental design, analysis, interpretation, manuscript preparation and the decision to submit should be stated. If the work received no specific funding, this should also be declared.
Data, Code and Model Availability
Authors should include a Data Availability Statement describing where the data supporting the findings can be accessed or explaining why data cannot be shared.
Where possible and appropriate, authors should provide access to code, model weights, model cards, data cards, configuration files, prompts, evaluation scripts, annotation guidelines, synthetic data generation procedures or other research artifacts. Restrictions due to privacy, ethics, legal, commercial, intellectual property, licensing, cybersecurity, safety or misuse concerns should be clearly explained.
Use of AI and AI-assisted Technologies
AI tools and large language models cannot be listed as authors. Authors remain responsible for the accuracy, originality and integrity of all submitted content.
Use of generative AI or AI-assisted technologies beyond basic language editing should be disclosed according to the journal’s AI policy. Authors should identify the tool used, describe how it was used and verify all AI-assisted content, including references, code, factual claims, generated text and mathematical statements.
Acknowledgements
Individuals or organizations that contributed to the work but do not meet authorship criteria should be acknowledged with their permission.
Before submitting a manuscript, authors should ensure that the following items are complete:
the manuscript fits the journal’s Aims and Scope;
the correct article type has been selected;
the manuscript follows the relevant structure and formatting requirements;
the title page includes complete author, affiliation and corresponding author information;
all authors have approved the submitted version;
authorship, author order and corresponding author details have been confirmed;
the abstract and keywords are included;
all tables, figures, screenshots and supplementary materials are cited in the text and uploaded in the required format;
figure legends and table titles are complete;
supplementary materials are cited and uploaded where applicable;
references are accurate and complete;
model, dataset, software, version, intended use and evaluation setting are described where applicable;
training, validation and test data are described clearly where applicable;
performance metrics, uncertainty, limitations and bias or fairness considerations are reported where applicable;
ethics approval, informed consent, data governance and trial registration information are provided where applicable;
author contributions, competing interests, funding and data availability statements are included;
data, code, model or artifact availability is described where applicable;
sponsor, funder, data provider, technology provider or platform provider involvement is disclosed where applicable;
AI-use disclosure is included where applicable;
reporting checklists and flow diagrams are uploaded for relevant study types;
permission has been obtained for third-party material;
the cover letter has been prepared;
all required files have been uploaded to the submission system.
Manuscripts should be submitted through the journal’s online submission system. Submissions sent by email are generally not considered unless specifically requested by the Editorial Office.
During submission, authors may be asked to provide:
manuscript title and abstract;
author names, affiliations and email addresses;
corresponding author details;
article type;
keywords;
cover letter;
manuscript file;
figures, tables and supplementary files;
required statements and declarations;
reporting checklists;
suggested or opposed reviewers, where permitted;
funding and competing interest information.
The cover letter should briefly explain the importance of the work, its relevance to the journal and why it may be of interest to the journal’s readership. Authors should also disclose any related manuscripts, previous submissions, preprints, conference papers, workshop papers, technical reports, dataset documentation, model release notes, software documentation or overlapping work.
Authors may suggest potential reviewers where permitted by the submission system. Reviewer selection remains at the editors’ discretion. Suggested reviewers must be independent and free from conflicts of interest.
After submission, the corresponding author will receive confirmation from the submission system. The Editorial Office may contact the corresponding author if required information is missing or if files need correction before editorial assessment.
If revision is invited, authors should submit the revised manuscript within the deadline stated in the decision letter. Extensions may be requested from the Editorial Office where necessary.
A revised submission should usually include:
a clean version of the revised manuscript;
a marked-up or tracked-changes version showing revisions;
a detailed point-by-point response to all editor and reviewer comments;
revised figures, tables or supplementary materials, where applicable;
updated declarations or statements, if any changes have been made.
Authors should respond to each comment respectfully and clearly. Where authors disagree with a reviewer or editor comment, they should provide a reasoned explanation supported by evidence where appropriate.
If new experiments, additional validation, updated models, new analyses or revised datasets are added during revision, authors should update the Methods, Results, limitations, data availability statement and any relevant reporting checklist accordingly.
Submission of a revised manuscript does not guarantee acceptance. Revised manuscripts may be returned to reviewers, assessed by the editors or declined if concerns have not been adequately addressed.
After acceptance, authors may be required to complete a publishing agreement, confirm copyright and licensing arrangements, select open access options where applicable and pay any applicable publication charges.
The corresponding author will receive proofs for review before publication. Proof corrections should be limited to typographical errors, formatting issues and factual corrections. Substantial changes to the content, authorship, data, analysis or conclusions after acceptance may require editorial approval.
Accepted articles may be published online before assignment to an issue, where applicable. The final published article will be assigned a DOI.
After publication, authors may share their article according to the journal’s article sharing and self-archiving policies. Requests for corrections, retractions, expressions of concern or other post-publication updates will be handled according to the journal’s relevant policies.
