We invite original research full papers and extended abstracts for the workshop “Building Robust, Trustworthy, and Safe AI for Healthcare”.
The workshop examines how AI systems can be evaluated and used responsibly across diverse patient populations, clinical settings, and public health services.
We welcome methodological research, empirical evaluations, deployment case studies, and position papers grounded in evidence or practical experience. Negative results, analyses of failures, and lessons from real-world use are encouraged. Applications may involve clinical records, medical images, and other health data, spanning predictive models, foundation models, and generative AI.
Click here for full information on the workshop.
!! Deadlines !!
| Workshop Call for Papers (CFP) – Submission deadline | 23 December 2026 |
| Workshop Call for Papers (CFP) – Notification deadline | 28 January 2027 |
| Camera-ready deadline | 7 April 2027 |
Themes of interest
- Robustness and generalization across hospitals, populations, devices, and data sources, including distribution shift and incomplete or noisy data.
- Trustworthy evaluation: external and multicenter validation, clinically meaningful benchmarks, reproducibility, and prevention of data leakage.
- Uncertainty estimation, calibration, out-of-distribution detection, failure identification, and deferral to healthcare professionals.
- Explainability, transparency, and communication of model capabilities and limitations.
- Fairness, subgroup performance, bias mitigation, and health equity.
- Safety, security, privacy-preserving learning, adversarial risks, and assessment of potential harms.
- Human-AI collaboration, workflow integration, post-deployment monitoring, and responsible model updating.
- Ethical, regulatory, and governance approaches to accountable healthcare AI.
- Reliability and safety of foundation models, generative AI, and multimodal healthcare systems.
Submission guidelines
We encourage interested author(s) to submit either a full paper or an extended abstract. The author(s) of an accepted paper/ abstract, which is presented in-person at DPH 2027 in London, will be included in the conference proceedings – extended abstracts will be included in Frontiers proceedings, and full papers will be submitted for possible inclusion into IEEE Xplore.
For submitting a Full Paper, please follow the instructions:
- All submissions must be not simultaneously submitted to another journal or conference.
- Scope: Original empirical, methodological or sufficiently developed conceptual research.
- The papers must range between 4‑8 pages, including references and any appendices.
- Papers should conform to the IEEE double column format: https://www.ieee.org/conferences/publishing/templates.html
- Papers must be anonymised: the names and affiliations removed, and any references to previous works either substituted with “anonymised” or rephrased.
- All submissions must be submitted in PDF format using the Easychair submission system. Click on the submission button below and choose the “DPH2027 Workshop – Building Robust, Trustworthy, and Safe AI for Healthcare” track.
- Please paste your Abstract in the Abstract field in the Easychair system for ease of processing and review allocation.
NOTE that:
- Full Paper submissions will be reviewed by three reviewers from the PC and SPC committees and assessed in terms of their quality, relevance to the conference, originality, significance, correctness, and clarity.
- All accepted papers will have their own DOI.
- Accepted papers will need to incorporate reviewers’ comments and re-submit their paper by the Camera-ready deadline. Instructions will be provided.
- Papers accepted for publication must have at least one author registered and must be presented in person at the conference in order to be considered for publication by IEEE Xplore. Papers from no-show authors will be removed from the proceedings.
For submitting an Extended Abstract, please follow the instructions:
- Please format your submission using Times New Roman, 12 pt font size, with justified alignment. The submission must not exceed 1000 words, excluding the reference list. References are compulsory and should be included in the submission. A maximum of two figures or tables may be included.
- Scope: Research results, educational interventions, reviews or well-specified work in progress.
- Extended Abstracts could follow the standard structure: Introduction, Background, Methods / Datasets, Results, and Conclusion. Alternatively, computer science and engineering abstracts can follow alternative structure such as Introduction, Background, Framework / Architecture, Implementation, Results (if available), Novelty and Challenges and Conclusion.
- Submissions must be anonymised: the names and affiliations removed, and any references to previous works either substituted with “anonymised” or rephrased.
- All submissions must be submitted in PDF format using the Easychair submission system. Click on the submission button below and choose the “DPH2027 Workshop – Building Robust, Trustworthy, and Safe AI for Healthcare” track.
- Paste your first section (Aims/Abstract/Overview) to the Abstract field in the Easychair system for ease of processing and review allocation.
NOTE that:
- Abstract submissions will be reviewed by two-three reviewers from the PC and SPC committees and assessed in terms of their quality, relevance to the conference, originality, significance, correctness, and clarity.
- Accepted Extended Abstracts will have to incorporate reviewers’ comments and re-submit their abstract by the Camera-ready deadline. Instructions will be provided.
Review and selection
Submissions will be assessed for relevance, originality, soundness, clarity, and the evidence supporting their claims. Case studies and position papers will also be assessed on their practical insights and the strength of their arguments. Conflicts of interest will be managed in reviewer assignment and acceptance decisions.
Presentation and publication
The workshop is in person in London. Accepted contributions are provisionally planned for poster presentation. Display and discussion arrangements will be finalised with DPH.
Contact
All questions about submissions and the workshop should be emailed to:
- Laya Rafiee Sevyeri – laya.rafieesevyeri@mail.mcgill.ca
- Steven Fernandes – StevenFernandes@creighton.edu
General questions about the conference can be sent to support@dphconf.org.
SUBMISSION
Submission is via EasyChair.