Introduction
Clinical decisions rarely rest on a single image and they often combine imaging with vital signs, laboratory results, medication histories, and free-text notes that change over hours and days. Multimodal foundation models represent the heterogeneous modalities of image, text and structured records jointly, while agentic frameworks offer the control layer that decides what to look at next. Together, they can now plan, call tools, retrieve evidence, and reason across these diverse modalities. This workshop brings together researchers in medical imaging, clinical machine learning, and agentic AI to examine how multimodal agents should be designed, grounded, evaluated, and deployed for clinical decision support.
Call for Papers
Important Dates
- Submission Portal Opens: December 1, 2026
- Submission Deadline: January 26, 2027, 23:59 AOE
Set by ISBI 2027 for all workshops. No extensions. - Review Period: January 27 – February 16, 2027
- Decisions: February 23, 2027
- Camera Ready: March 9, 2027, 23:59 AOE
- Workshop: May 25, 2027, Lausanne
Guidelines
- Paper submission will be hosted on the submission portal linked above once it opens.
- We welcome papers up to 4 pages in the IEEE ISBI format, plus one page for references.
- The paper should be anonymized and uploaded as a single PDF.
- Posting papers on preprint servers like arXiv is permitted.
- The LaTeX template will be linked here when the portal opens.
- Accepted papers will be presented as orals or posters. At least one author must register for ISBI 2027 and present in person.
- This is a non-archival workshop. No submission will be indexed nor have archival proceedings.
- Accepted papers will appear on the workshop website.
- We accept submissions that are under review at other venues (e.g., MICCAI 2027), as long as this does not violate the dual-submission or anonymity policy of the other venue.
- The review process will be single-blind.
Topics of Interest
Foundation models and vision-language models have advanced medical image analysis quickly. Clinical decisions, however, are multimodal and longitudinal: an ICU clinician reads a chest radiograph alongside ventilator settings, lab trends, and nursing notes accumulated across a stay. Agentic systems that plan, call specialized tools such as segmentation or retrieval models, and combine the evidence are a natural fit for this setting. How should agents ground their claims in image regions? How should they handle patient histories that exceed context limits? How do we evaluate the reasoning and not only the final answer? How do these systems fail across hospitals and scanners? We welcome works that address these questions across imaging, clinical machine learning, and agentic AI, including but not limited to:
- Multimodal fusion of imaging with EHR, physiological time series, and clinical text
- Agent architectures, tool use, and multi-agent systems for diagnosis, triage, and prognosis
- Long-context and memory mechanisms for patient trajectories
- Spatial, anatomical, and visual grounding of agent outputs
- Benchmarks and evaluation protocols that reflect clinical workflows
- Robustness, uncertainty, and generalization across sites and scanners
- Human-AI interaction, safety, and clinical translation
Schedule
May 25, 2027 (CEST), 09:30–15:30.
| 09:30–09:40 | Opening remarks and workshop overview (10 min) |
| 09:40–10:10 | Keynote: Multimodal AI agents for clinical decision support(30 min) |
| 10:10–10:40 | Speaker 1: Assistive robotic agents in healthcare(30 min) |
| 10:40–10:55 | Joint keynote Q&A and discussion (15 min) |
| 10:55–11:10 | Coffee break(15 min) |
| 11:10–12:10 | Oral session: three to four selected short papers (60 min) |
| 12:10–13:10 | Lunch break (60 min) |
| 13:10–13:50 | Poster session and networking(40 min) |
| 13:50–14:10 | Speaker 2: Decision-making in critical care and what clinicians need from AI(20 min) |
| 14:10–14:25 | Speaker 3 Q&A and transition(35 min) |
| 14:25–15:00 | Panel discussion: Evaluating and deploying multimodal agents safely(35 min) |
| 15:00–15:25 | Coffee + Open discussion, roundtable(25 min) |
| 15:25–15:30 | Best paper award and closing remarks (5 min) |
Organizers

Umaima Rahman
PostDoctoral Associate
New York University AD

Farah E. Shamout
Assistant Professor
New York University AD

Dwarikanath Mahapatra
Assistant Professor
Khalifa University

mohammad alsharid
Assistant Professor
Khalifa University

Congbo Ma
PostDoctoral Associate
New York University AD

Julián Lechuga López
PhD Student
New York University AD

Baraa Al Jorf
PhD Student
New York University AD

Saeed Ahmad
PhD Student
New York University AD

Chaimae Abouzahir
Research Assistant
New York University AD