Clinical workflows are increasingly information-intensive, requiring the joint interpretation of heterogeneous data sources, including medical images, electronic health records (EHRs), clinical notes, and biomedical knowledge. While these complementary sources provide a richer understanding of patient conditions, they also introduce significant challenges in information retrieval, integration, fusion, and clinical reasoning.

Recent advances in multimodal foundation models, vision-language models, and retrieval-augmented generation have created new opportunities for AI-based clinical decision support. However, important challenges remain, including cross-modal retrieval, representation alignment, multimodal fusion, and the development of robust, fair, interpretable, and privacy-preserving systems for high-stakes clinical environments.

The workshop Clinical-MIRF: Multimodal Information Retrieval and Fusion in Clinical Workflows addresses these challenges by positioning clinical AI as an information retrieval and knowledge integration problem. It aims to bring together researchers from information retrieval, knowledge management, multimodal learning, biomedical Natural Language Processing (NLP), medical imaging, EHR analytics, and clinical AI to discuss methods, datasets, evaluation protocols, and deployment challenges for multimodal clinical information systems.

The main objectives of the workshop are:

• Establish multimodal retrieval and fusion for clinical workflows as an emerging research area.

• Foster collaboration across information retrieval, knowledge management, medical imaging, biomedical NLP, and clinical AI communities.

• Discuss challenges in retrieval, fusion, alignment, reasoning, evaluation, robustness, fairness, and privacy.

• Identify open datasets, benchmarks, evaluation protocols, and deployment barriers.

Topics include, but are not limited to:

• Multimodal information retrieval and cross-modal search

• Retrieval-augmented generation for healthcare

• Representation learning and modality alignment

• Vision-language and multimodal foundation models

• Multimodal fusion for clinical decision support

• Knowledge graphs and clinical knowledge integration

• Explainability, trustworthiness, robustness, and fairness in multimodal clinical AI

Submission Guidelines

All papers have to follow the template as indicated here: https://www.acm.org/publications/proceedings-template

Full research papers must not exceed 10 pages, including all figures, tables, and appendices, plus up to 2 additional pages exclusively for references.

Organizing Committees

• Michela Gravina, University of Naples Federico II, michela.gravina@unina.it

• Antonio Galli, University of Naples Federico II, antonio.galli@unina.it

• Antonino Ferraro, Pegaso University, antonino.ferraro@unipegaso.it

• Martina Iammarino, Pegaso University, martina.iammarino@unipegaso.it

• Valerio Guarrasi, UniCamillus-Saint Camillus International University of Health Sciences, valerio.guarrasi@unicamillus.org

• Angel Mario García-Pedrero, Universidad Politécnica de Madrid, angelmario.garcia@upm.es

• Consuelo Gonzalo-Martín, Universidad Politécnica de Madrid, consuelo.gonzalo@upm.es

• Vincenzo Moscato, University of Naples Federico II, vincenzo.moscato@unina.it

• Paolo Soda, Università Campus Bio-Medico di Roma, p.soda@unicampus.it

• Carlo Sansone, University of Naples Federico II, carlo.sansone@unina.it