Manuscript Template | Health Nexus: Digital Health and Medical AI

This template provides the recommended structure and formatting requirements for manuscripts submitted to Health Nexus: Digital Health and Medical AI. Authors should adapt the template to the selected article type while ensuring compliance with the journal’s Author Guidelines, reporting standards, research-ethics requirements, and double-anonymous peer-review procedures.

Authors must prepare and upload the following files separately:

Title Page — containing complete author and declaration information.
Anonymous Manuscript — containing no information that identifies the authors or their institutions.
Supplementary Files — including reporting checklists, protocols, appendices, data documentation, code documentation, figures, or other supporting materials where applicable.

The templates below indicate the required order and content of each file.

1. General Formatting Requirements

Submit the manuscript in an editable Microsoft Word format, preferably .docx.
Use a standard, readable font such as Times New Roman, Arial, or Calibri.
Use 11- or 12-point font for the main text.
Use double-line spacing throughout the manuscript, including the abstract, references, tables, and figure legends.
Use margins of approximately 2.5 cm on all sides.
Number all pages consecutively.
Apply continuous line numbering to the anonymous manuscript.
Use left-aligned text rather than full justification.
Use clear heading levels consistently.
Define abbreviations at first use.
Use internationally recognised scientific terminology and SI units where appropriate.
Do not use decorative formatting, coloured text, text boxes, or unnecessary page design elements.
Tables must remain editable and should not be inserted as images.
Figures should be submitted at publication-quality resolution.

The manuscript should be written in clear academic English. Authors are responsible for checking grammar, spelling, terminology, numerical consistency, references, and the accuracy of all scientific statements before submission.

2. File 1: Title Page Template

The title page is not normally provided to external reviewers. It must contain all author-identifying information and publication declarations.

Manuscript Title

[Insert the full manuscript title]

The title should be concise, informative, and accurately reflect the design, population, intervention, technology, or principal subject of the study. Avoid unexplained abbreviations, promotional language, and unsupported claims.

Short Running Title

[Insert a shortened title of no more than approximately 50 characters]

Article Type

[Select the appropriate article type, for example: Original Research Article, Review Article, Systematic Review and Meta-Analysis, Methodological Article, Technical and Software Article, Implementation and Evaluation Study, Protocol, Case Study, Perspective, Commentary, or Letter to the Editor]

Authors

[Full name of Author 1]¹, [Full name of Author 2]², [Full name of Author 3]^1,3

Authors’ names should be presented consistently and in the order agreed by all contributors.

Institutional Affiliations

¹ [Department, Institution, City, Country]

² [Department, Institution, City, Country]

³ [Department, Institution, City, Country]

Affiliations should identify the institution at which the relevant work was conducted. Honorifics, academic degrees, professional titles, and administrative positions should not be included in the author-name line.

ORCID Identifiers

[Author 1]: [ORCID]
[Author 2]: [ORCID]
[Author 3]: [ORCID]

Corresponding Author

Name: [Full name]

Institution: [Institutional affiliation]

Postal address: [Complete institutional address]

Email: [Institutional or professional email address]

ORCID: [ORCID identifier, where available]

Word Count and Manuscript Components

Main-text word count: [Insert number]
Abstract word count: [Insert number]
Number of references: [Insert number]
Number of tables: [Insert number]
Number of figures: [Insert number]
Number of supplementary files: [Insert number]

Author Contributions

Authors should describe individual contributions, preferably using the CRediT contributor-role taxonomy.

Suggested format:

[Author initials]: Conceptualisation, Methodology, Investigation, Formal Analysis, Writing – Original Draft.

[Author initials]: Software, Validation, Data Curation, Visualisation, Writing – Review and Editing.

[Author initials]: Supervision, Project Administration, Funding Acquisition, Writing – Review and Editing.

All authors should have reviewed and approved the final manuscript and agreed to be accountable for the integrity of the work.

Funding

Example where funding was received:

This work was supported by [name of funding organisation] under Grant [grant number].

Example where no specific funding was received:

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

The role of the funder in study design, data collection, analysis, interpretation, manuscript preparation, and the decision to submit should be disclosed.

Conflict of Interest

Example where no conflict exists:

The authors declare that they have no competing financial or non-financial interests relevant to the work reported in this article.

Example where a conflict exists:

[Author initials] has received [consultancy fees, research support, honoraria, stock ownership, patent income, or other interest] from [organisation]. The remaining authors declare no competing interests.

Acknowledgements

[Acknowledge individuals, institutions, technical support, language assistance, data-access support, or other contributions that do not meet the criteria for authorship.]

Authors should obtain permission from individuals named in the acknowledgements.

Ethics Approval

Example:

The study was approved by the [full name of ethics committee or institutional review board] under approval number [number]. The study was conducted in accordance with applicable institutional and national ethical standards.

Where ethical approval was not required:

Ethical approval was not required for this study because [provide a specific and appropriate explanation].

Informed Consent

Example:

Written informed consent was obtained from all participants before participation in the study.

Where consent was waived:

The requirement for informed consent was waived by the [name of ethics committee] because [reason].

Consent for Publication

Example:

Written informed consent for publication of identifiable information and images was obtained from the participant or the participant’s legally authorised representative.

Where not applicable:

Consent for publication was not applicable because the manuscript contains no identifiable personal information.

Clinical-Trial Registration

Registry: [Name of recognised registry]

Registration number: [Registration identifier]

Date of registration: [Date]

URL: [Registry record address]

Where the study is not a clinical trial, state:

Clinical-trial registration was not applicable.

Data Availability Statement

Select or adapt the statement most appropriate to the study.

Publicly available data:

The data supporting the findings of this study are available in [repository name] at [persistent identifier or repository address].

Data available on reasonable request:

The data supporting the findings of this study are available from the corresponding author on reasonable request, subject to applicable ethical, legal, institutional, and privacy restrictions.

Restricted clinical or personal data:

The data are not publicly available because they contain potentially identifiable or sensitive health information. Access may be considered upon reasonable request and approval by the relevant institution or ethics committee.

No new data:

No new datasets were generated or analysed during this study.

Code and Model Availability

Example:

The analysis code and model documentation are available in [repository] at [persistent identifier].

Restricted availability:

The source code or model components cannot be made publicly available because of [commercial, security, licensing, privacy, or institutional restriction]. Relevant documentation may be made available upon reasonable request where legally and technically permissible.

Generative AI and AI-Assisted Technology Declaration

Where no generative AI was used:

The authors declare that no generative artificial intelligence or AI-assisted technology was used to generate scientific content, analyse data, produce images, or make substantive intellectual contributions to this manuscript.

Where AI-assisted tools were used:

During the preparation of this manuscript, the authors used [tool name, provider, and version] for [language editing, translation, coding assistance, data analysis, image preparation, or another specified purpose]. All outputs were reviewed, verified, and revised by the authors, who accept full responsibility for the accuracy, originality, integrity, and final content of the manuscript.

Prior Dissemination

Authors must disclose whether the work has previously appeared in a preprint, conference abstract, thesis, dissertation, institutional report, working paper, repository, or other public form.

Example:

An earlier version of this manuscript was deposited as a preprint at [repository] under identifier [identifier]. The manuscript has been substantially revised for journal submission.

3. File 2: Anonymous Manuscript Template

The anonymous manuscript is provided to peer reviewers and must not contain information that directly or indirectly identifies the authors.

Authors must remove:

Author names and affiliations
Corresponding-author information
ORCID identifiers
Acknowledgements identifying individuals or institutions
Named grant holders where this would disclose authorship
Institution-specific ethics information where disclosure would identify the authors
Institutional logos, headers, footers, or file names
Identifying document properties and tracked-change metadata

Where information must be temporarily concealed, authors may use neutral placeholders such as:

[Institution blinded for peer review]
[Ethics committee details blinded for peer review]
[Funding information provided on the title page]
[Acknowledgements provided on the title page]

Manuscript Title

[Insert the full manuscript title]

The anonymous manuscript should begin with the manuscript title. Do not include author names or affiliations beneath the title.

Abstract

Original research, clinical studies, implementation studies, diagnostic studies, and algorithm-development or validation studies should normally use a structured abstract.

Background: [Briefly explain the clinical, technological, public-health, or scientific context.]

Objective: [State the principal objective, research question, or hypothesis.]

Methods: [Summarise the study design, setting, participants or datasets, technology or model, comparison methods, outcomes, and analytical approach.]

Results: [Report the principal numerical or qualitative findings. Include sample size, performance estimates, effect estimates, confidence intervals, and relevant validation results where applicable.]

Conclusions: [State conclusions supported by the results without exaggerating clinical readiness, effectiveness, or generalisability.]

Trial Registration: [Registry and registration number, where applicable.]

Reviews, Perspectives, Commentaries, and other non-research article types may use an unstructured abstract where appropriate.

Keywords

[Keyword 1]; [Keyword 2]; [Keyword 3]; [Keyword 4]; [Keyword 5]

Provide four to eight specific and searchable keywords. Medical Subject Headings should be used where appropriate.

4. Main Text Template for Original Research Articles

1. Introduction

The Introduction should:

Explain the relevant clinical, public-health, technical, or policy context.
Summarise the current state of knowledge.
Identify the specific evidence gap or unresolved problem.
Explain why the study is necessary.
State the study objective, research question, and hypothesis where applicable.

The Introduction should remain focused. It should not provide an exhaustive literature review or report study results.

Suggested final paragraph:

Accordingly, this study aimed to [state the principal objective]. We hypothesised that [state hypothesis, where applicable].

2. Methods

The Methods section must provide sufficient detail for critical evaluation and, where reasonably possible, replication.

2.1 Study Design

[Identify the study design, such as randomised trial, prospective cohort, retrospective cohort, cross-sectional study, diagnostic-accuracy study, qualitative study, mixed-methods study, algorithm-development study, external-validation study, or implementation study.]

State the applicable reporting guideline.

Example:

This study was reported in accordance with the [name of guideline] statement.

2.2 Study Setting

[Describe the clinical, institutional, community, laboratory, virtual, or digital setting.]

Report relevant geographic locations, healthcare-system characteristics, study dates, and the period of data collection.

2.3 Participants, Records, or Data Sources

Describe:

Eligibility criteria
Recruitment or sampling procedures
Sources of health records or datasets
Inclusion and exclusion procedures
Participant or record selection
Sample size
Relevant demographic and clinical characteristics

For secondary datasets, provide the dataset name, custodian, collection period, geographic origin, inclusion criteria, data-access conditions, and any known limitations.

2.4 Intervention, Technology, or System

Where applicable, describe:

The digital-health intervention or AI system
The intended users
The intended clinical or public-health purpose
The hardware and software environment
The model or system version
The workflow in which the system was used
The comparator or standard of care
The degree of human oversight

2.5 Outcomes and Variables

Define all primary and secondary outcomes, predictors, exposures, reference standards, endpoints, and measurement procedures.

State whether outcomes and analyses were prespecified.

2.6 Data Collection and Preprocessing

Describe:

Data-acquisition procedures
Data-cleaning procedures
Handling of missing data
Data labelling and annotation
Quality-control procedures
Normalisation or transformation
Feature selection or feature engineering
Image, signal, text, or record preprocessing
De-identification or privacy-protection procedures

2.7 Artificial-Intelligence or Machine-Learning Model

For studies involving AI or machine learning, report:

Model type and architecture
Software libraries and versions
Training procedures
Hyperparameters
Optimisation procedures
Stopping criteria
Model-selection procedures
Training, validation, and test-set separation
Cross-validation procedures
Measures used to prevent data leakage
External or temporal validation
Comparator models
Explainability or interpretability methods

Authors should state clearly whether the study concerns model development, internal validation, external validation, clinical evaluation, implementation, or post-deployment monitoring.

2.8 Generative AI or Large Language Model Procedures

Where applicable, report:

Provider and model name
Model version
Date of access
System prompts and user prompts
Prompt-development procedures
Generation settings or sampling parameters
Number of repeated queries
Retrieval-augmented generation or external tools
Human-review procedures
Evaluation criteria
Assessment of hallucination, bias, safety, and factual accuracy
Handling of confidential or protected health information

2.9 Sample-Size Determination

Describe the sample-size calculation, statistical assumptions, expected effect size, power, significance threshold, outcome prevalence, model complexity, or other applicable basis for determining the sample size.

For machine-learning studies, authors should explain how the sample size was considered in relation to the number of predictors, outcome events, class imbalance, and model complexity.

2.10 Statistical Analysis

Describe:

Statistical tests and models
Effect measures
Confidence intervals
Significance thresholds
Adjustment variables
Handling of missing data
Multiple-comparison procedures
Sensitivity and subgroup analyses
Model diagnostics
Software and version numbers

For predictive or diagnostic models, include relevant measures of discrimination, calibration, classification performance, clinical utility, and uncertainty.

2.11 Ethics and Governance

State the applicable ethical approval, informed-consent arrangements, privacy protections, data-governance procedures, security controls, and legal basis for using the relevant data.

During anonymous review, use a neutral placeholder where necessary:

This study was approved by [ethics committee details blinded for peer review] under approval number [blinded].

3. Results

The Results section should present findings in a logical order without unnecessary interpretation.

3.1 Study Population or Dataset

Report:

The number assessed for eligibility
The number included and excluded
Reasons for exclusion
Participant or dataset characteristics
Missing data
Follow-up where relevant

Use a flow diagram where required by the relevant reporting guideline.

3.2 Primary Findings

Report the principal findings with appropriate numerical estimates, uncertainty measures, confidence intervals, and exact denominators.

3.3 Model Performance

Where applicable, report:

Performance in training, validation, and independent test datasets
Discrimination measures
Calibration measures
Sensitivity and specificity
Positive and negative predictive values
Precision, recall, and F1 score
Area under the receiver-operating-characteristic curve
Decision-curve or clinical-utility measures
Error analysis
Failure cases
Subgroup performance

Performance measures should be selected according to the clinical purpose, data distribution, outcome prevalence, and study design.

3.4 Secondary and Sensitivity Analyses

Clearly distinguish prespecified analyses from exploratory analyses.

3.5 Adverse Events, Errors, and Safety Findings

Where relevant, report system failures, unexpected outputs, adverse events, misclassifications, unsafe recommendations, workflow disruptions, privacy incidents, or other clinically important errors.

4. Discussion

The Discussion should normally include:

4.1 Principal Findings

Summarise the main findings without repeating all numerical results.

4.2 Comparison with Previous Research

Interpret the findings in relation to relevant prior studies, established standards, and competing technologies or approaches.

4.3 Clinical, Technical, or Public-Health Implications

Explain the potential significance of the findings while distinguishing clearly between demonstrated evidence and future possibilities.

4.4 Strengths and Limitations

Discuss relevant limitations, including:

Study-design limitations
Selection bias
Measurement bias
Dataset limitations
Class imbalance
Missing data
Limited external validation
Geographic or institutional limitations
Algorithmic bias
Restricted model transparency
Reproducibility limitations
Workflow or implementation constraints
Changes in model versions or technology over time

4.5 Future Research

Identify the additional validation, clinical evaluation, implementation research, regulatory assessment, or long-term monitoring required.

5. Conclusions

The Conclusions should be concise and directly supported by the results. Authors should avoid claiming clinical effectiveness, safety, superiority, generalisability, or implementation readiness unless these conclusions are supported by the study design and evidence.

5. Declarations in the Anonymous Manuscript

The anonymous manuscript should include declaration headings where required, but identifying details should be removed or replaced with neutral placeholders.

Ethics Approval and Consent to Participate

[Provide the required statement in anonymised form.]

Consent for Publication

[Provide the required statement or state that it is not applicable.]

Data Availability

[Provide the data-availability statement without identifying the authors.]

Code and Model Availability

[Provide the relevant availability statement.]

Competing Interests

[Provide the declaration without unnecessary author-identifying details.]

Funding

[State whether funding was received. Funding details that reveal author identity may be provided only on the title page during anonymous review.]

Generative AI Declaration

[Provide the applicable declaration.]

Acknowledgements

[Acknowledgements provided on the title page and withheld during anonymous peer review.]

6. References

All manuscripts submitted to Health Nexus: Digital Health and Medical AI must use the PSG Author–Date Citation Format, the official citation standard of Panorama Scholarly Group.

PSG Format uses parenthetical author–date citations in the main text and a corresponding reference list. Authors must ensure that every source cited in the manuscript appears in the reference list and that every entry in the reference list is cited in the manuscript.

6.1 In-Text Citations

In-text citations should include the author’s surname and year of publication. A comma is not placed between the author’s surname and the year.

Citation Type	PSG Format
One author	(Smith 2024)
Two authors	(Smith and Lee 2024)
Three authors	(Smith, Lee, and Wang 2024)
Four or more authors	(Smith et al. 2024)
Specific page	(Smith 2024, 25)
Page range	(Smith 2024, 25–27)
Multiple sources	(Chen 2021; Kim 2022; Smith 2024)
No publication date	(Smith n.d.)
Same author and same year	(Smith 2024a, 2024b)

When an author’s name forms part of the sentence, only the year and relevant page number should appear in parentheses.

Example:

Smith (2024) argued that clinical artificial-intelligence systems require continuous post-deployment evaluation.

Smith (2024, 25–27) further identified data drift as a major source of clinical risk.

Direct quotations must include a page number or another appropriate location indicator. Authors should use direct quotations sparingly and should normally paraphrase and critically analyse source material.

6.2 General Reference-List Rules

The year of publication appears immediately after the author information.
The first author’s name is inverted: surname first, followed by the given name.
The names of subsequent authors appear in natural order: given name followed by surname.
Use and before the final author’s name; do not use an ampersand.
English article and chapter titles should appear in typographic quotation marks: “ ”.
The period following an article or chapter title should appear inside the closing quotation mark.
Journal names and book titles should be italicised.
Volume and issue information should follow the form: 12, no. 2.
Page ranges must use an en dash, for example: 45–63.
Where an article has an article number rather than page numbers, use the form: Article 108.
DOIs must be presented as complete DOI links beginning with https://doi.org/.
Do not place a full stop after a DOI or URL.
Authors must verify all names, titles, dates, volume numbers, issue numbers, page ranges, article numbers, DOIs, and URLs against the original source.
Retracted publications must not be cited as reliable evidence unless the retraction itself is the subject of discussion.
Fabricated, unverifiable, incomplete, or AI-generated references are prohibited.

6.3 Journal Article

Format:

Last, First, First Last, and First Last. Year. “Article Title.” Journal Name volume, no. issue: page range. DOI

Example:

Smith, John A., Helen K. Lee, and Ming Wang. 2024. “Clinical Validation of Artificial Intelligence in Remote Patient Monitoring.” Journal of Digital Medicine 12, no. 2: 45–63. https://doi.org/10.xxxx/xxxxx

In-text citation: (Smith, Lee, and Wang 2024)

6.4 Journal Article with an Article Number

Example:

Chen, Li, and Sungho Park. 2025. “Evaluation of a Machine-Learning System for Hospital Risk Prediction.” Digital Health Research 9, no. 1: Article 108. https://doi.org/10.xxxx/xxxxx

In-text citation: (Chen and Park 2025)

6.5 Online-First Journal Article

Example:

Garcia, Elena, and David Brown. 2026. “Generative Artificial Intelligence in Clinical Documentation.” Medical Informatics Review. Published online January 15, 2026. https://doi.org/10.xxxx/xxxxx

In-text citation: (Garcia and Brown 2026)

6.6 Book

Format:

Last, First. Year. Book Title: Subtitle. Place of publication: Publisher.

Example:

Topol, Eric. 2019. Deep Medicine: How Artificial Intelligence Can Make Healthcare Human Again. New York: Basic Books.

In-text citation: (Topol 2019)

6.7 Chapter in an Edited Book

Example:

Lee, Hyun K. 2024. “Artificial Intelligence and Clinical Decision-Making.” In Digital Transformation in Healthcare, edited by John Smith and Robert Brown, 55–78. Singapore: Springer. https://doi.org/10.xxxx/xxxxx

In-text citation: (Lee 2024, 60)

6.8 Webpage

Format:

Author or Organisation. Year. “Title of Webpage.” Accessed Month Day, Year. URL

Example:

World Health Organization. 2025. “Digital Health and Artificial Intelligence.” Accessed August 5, 2026. https://example.org/digital-health

In-text citation: (World Health Organization 2025)

Where no publication or revision date can be identified, use n.d. and retain the access date.

6.9 Government or Institutional Report

Example:

Ministry of Health. 2025. National Strategy for Digital Health Development. Seoul: Ministry of Health. https://example.gov/digital-health-report

In-text citation: (Ministry of Health 2025)

6.10 Conference Paper

Example:

Brown, David, and Maria Garcia. 2025. “Clinical Evaluation of a Large Language Model for Patient Triage.” Paper presented at the International Conference on Medical Artificial Intelligence, Berlin, Germany, September 12–14.

In-text citation: (Brown and Garcia 2025)

6.11 Thesis or Dissertation

Example:

Kim, Minsoo. 2024. “Algorithmic Bias in Clinical Decision-Support Systems.” PhD diss., Seoul National University.

In-text citation: (Kim 2024)

6.12 Preprint

Preprints must be identified explicitly as preprints.

Example:

Wilson, James, and Li Chen. 2026. “Evaluating Hallucination Risk in Medical Large Language Models.” Preprint, medRxiv. https://doi.org/10.xxxx/xxxxx

In-text citation: (Wilson and Chen 2026)

6.13 Dataset

Example:

Lee, Sungho. 2025. Clinical Dataset for Remote Patient Monitoring Research. Data set. Zenodo. https://doi.org/10.xxxx/xxxxx

In-text citation: (Lee 2025)

Dataset references should identify the creator, year, dataset title, resource type, repository, version where applicable, and persistent identifier.

6.14 Software or GitHub Repository

Example:

Chen, Li. 2025. Clinical AI Validation Toolkit. Version 2.1. Source code. GitHub. https://github.com/example/clinical-ai-toolkit

In-text citation: (Chen 2025)

Where available, authors should identify the software version, release date, repository, and DOI or archived release identifier.

6.15 Artificial-Intelligence Tool

Example:

OpenAI. 2026. ChatGPT. Large language model. https://chat.openai.com/

In-text citation: (OpenAI 2026)

Citing an AI tool does not replace the separate requirement to disclose how the tool was used. Authors must also provide an AI-use declaration describing the model, version, purpose, affected stages of the work, and procedures used to verify its outputs.

6.16 Chinese-Language Sources

For Chinese-language sources, author names should be romanised in Pinyin. The original Chinese title should be retained and followed by an English translation in square brackets.

Example:

Wang, Ming, and Hua Li. 2025. “人工智能辅助诊断的临床应用与伦理风险 [Clinical Applications and Ethical Risks of AI-Assisted Diagnosis].” 医学信息学杂志 [Journal of Medical Informatics] 18, no. 2: 45–58. https://doi.org/10.xxxx/xxxxx

In-text citation: (Wang and Li 2025)

6.17 Korean-Language Sources

The original Korean title should be retained and followed by an English translation in square brackets.

Example:

Kim, Minsoo. 2025. “의료 인공지능의 임상적 검증과 책임성 [Clinical Validation and Accountability of Medical Artificial Intelligence].” 의료정보학연구 [Journal of Medical Informatics Research] 20, no. 1: 101–125. https://doi.org/10.xxxx/xxxxx

In-text citation: (Kim 2025)

6.18 Multiple Works by the Same Author in the Same Year

When an author or author group has more than one work published in the same year, lowercase letters should be added to the year and used consistently in both the text and reference list.

In-text citation:

(Smith 2024a, 2024b)

Reference-list examples:

Smith, John. 2024a. “Artificial Intelligence in Clinical Diagnosis.” Digital Medicine Review 8, no. 1: 1–15. https://doi.org/10.xxxx/xxxxx

Smith, John. 2024b. “Governance of Medical Artificial Intelligence.” Health Technology and Society 6, no. 3: 90–108. https://doi.org/10.xxxx/xxxxx

6.19 Reference Integrity

Authors are responsible for the completeness and accuracy of all references. Before submission, authors should:

Open and verify each DOI.
Confirm that each cited work exists and supports the statement for which it is cited.
Check author names, article titles, journal titles, publication years, volume and issue numbers, page ranges, and article numbers.
Identify preprints, datasets, software, and AI tools using the correct resource type.
Remove duplicate references.
Confirm that no retracted source is being presented as valid evidence.
Check all references generated or reformatted with reference-management or AI-assisted tools against the original sources.

Authors may use the official PSG Citation Generator provided by the Panorama Open Scholarly Index to assist with citation formatting. Automated output must still be checked manually for completeness, accuracy, capitalisation, multilingual metadata, and resource type.

7. Tables

Tables should be numbered consecutively in the order in which they are cited.

Example:

Table 1. Baseline characteristics of the study population

Characteristic	Development cohort	Validation cohort	P value
Number of participants	[n]	[n]	—
Age, mean (SD)	[value]	[value]	[value]
Female, n (%)	[value]	[value]	[value]

Abbreviations: SD, standard deviation.

Each table should:

Have a concise and informative title.
Be understandable without extensive reference to the main text.
Define all abbreviations in a footnote.
Identify statistical tests where relevant.
Specify whether values are numbers, percentages, means, medians, standard deviations, interquartile ranges, or confidence intervals.
Avoid unnecessary duplication of information presented in figures or text.

8. Figures and Figure Legends

Figures should be numbered consecutively in the order in which they are cited.

Example:

Figure 1. Study design and model-development workflow. The figure illustrates participant selection, data preprocessing, model training, internal validation, external validation, and final evaluation. AI, artificial intelligence; EHR, electronic health record.

Figure legends should:

Explain the figure clearly.
Define all abbreviations and symbols.
Identify error bars, confidence intervals, or statistical indicators.
State whether an image has been adapted from another source.
Identify any AI-generated or AI-modified visual material where its use is methodologically justified.

Figures should not contain identifiable patient information unless written consent for publication has been obtained.

9. Supplementary Material

Supplementary files should be labelled clearly and cited in the main manuscript.

Examples include:

Supplementary Appendix 1: Full search strategy
Supplementary Table 1: Additional subgroup analysis
Supplementary Figure 1: Calibration plots
Supplementary Methods: Detailed model architecture
Supplementary Code: Analysis scripts
Supplementary Prompts: Large language model prompts and system instructions
Reporting Checklist: CONSORT-AI, SPIRIT-AI, PRISMA, TRIPOD+AI, STARD-AI, DECIDE-AI, CLAIM, or another applicable checklist

Supplementary material must be accurate, appropriately anonymised, legally shareable, and consistent with the main manuscript.

10. Template for Review Articles

Review Articles may use the following structure:

Title
Abstract
Keywords
Introduction
Review objective or guiding questions
Search or literature-identification approach, where applicable
Thematic or conceptual sections
Critical analysis of the evidence
Methodological limitations of the literature
Implications for research, clinical practice, policy, or technology development
Future research priorities
Conclusions
Declarations
References

A narrative review should not be presented as a systematic review unless it follows a reproducible systematic method and the appropriate reporting guideline.

11. Template for Systematic Reviews and Meta-Analyses

Systematic Reviews and Meta-Analyses should normally include:

Structured abstract
Introduction
Research question and objectives
Protocol and registration
Eligibility criteria
Information sources
Complete search strategy
Study-selection process
Data-extraction procedures
Risk-of-bias assessment
Effect measures
Synthesis methods
Assessment of heterogeneity
Reporting-bias assessment
Certainty-of-evidence assessment, where applicable
Results
Study-selection flow diagram
Characteristics of included studies
Risk-of-bias findings
Synthesis or meta-analysis results
Discussion
Limitations
Conclusions

The complete search strategy and applicable PRISMA checklist should be uploaded as supplementary material.

12. Template for Technical and Software Articles

Technical and Software Articles should normally include:

Abstract
Keywords
Background and rationale
System objective and intended use
System architecture
Hardware and software requirements
Data inputs and outputs
Development methods
Security and privacy design
Interoperability standards
Validation and testing
Usability evaluation
Performance results
Comparison with existing systems
Implementation requirements
Limitations
Code and software availability
Conclusions

Authors should provide sufficient documentation to permit technical evaluation and, where possible, reproduction or reuse.

13. Template for Protocol Articles

Protocol Articles should normally include:

Structured abstract
Background
Study objectives
Study design
Study setting
Participants or datasets
Eligibility criteria
Intervention or technology
Outcome measures
Sample-size calculation
Recruitment or data-acquisition plan
Data-management plan
Statistical-analysis plan
AI or model-development procedures, where applicable
Ethics and consent
Data monitoring and safety procedures
Dissemination plan
Trial or protocol registration

Protocol manuscripts should clearly distinguish prespecified procedures from activities that will be determined after data collection begins.

14. Template for Perspectives and Commentaries

Perspectives and Commentaries may use a flexible structure but should normally include:

Brief abstract, where required
Introduction to the issue
Clear central argument or position
Evidence supporting the argument
Consideration of alternative perspectives
Clinical, ethical, technical, regulatory, or policy implications
Practical recommendations or future priorities
Conclusion

These articles should remain evidence-based and should distinguish personal interpretation from established evidence.

15. Final Pre-Submission Check

Before submission, authors should confirm that:

The title page and anonymous manuscript have been prepared as separate files.
The anonymous manuscript contains no identifying information.
The file metadata and tracked changes have been removed.
The correct article type has been selected.
The manuscript follows the relevant structure and reporting guideline.
The abstract and keywords comply with journal requirements.
All tables and figures are cited in numerical order.
References follow Vancouver style and have been verified.
Ethics approval, consent, trial registration, funding, conflicts of interest, data availability, and AI-use declarations have been included where applicable.
AI and machine-learning studies provide sufficient information concerning data, model development, validation, bias, performance, and reproducibility.
All authors have reviewed and approved the submitted version.

Manuscripts that substantially depart from this template may be returned for technical revision before editorial assessment. Authors may modify the section structure where justified by the article type or research methodology, but all information necessary for transparent scholarly evaluation must be retained.