Prof. Essam Rashed
Principal Investigator
AI • Medical Imaging • Multimodal Systems
Building trustworthy AI for medical imaging and clinical decision support.
We develop segmentation, detection, report-generation, and agentic AI systems that integrate imaging, clinical text, and external knowledge - emphasizing robustness, uncertainty, explainability, and real-world deployment.
Agentic AI
Multimodal LLM/VLM
Medical Image Segmentation
Federated Learning
Uncertainty & Safety
Regulatory-ready Evaluation
Research Areas
Segmentation & Detection
nnU-Net style pipelines, foundation-model adaptation, and clinically meaningful evaluation protocols.
Trustworthy Uncertainty & Safety
Calibration, abstention, and uncertainty-aware decision support for high-stakes deployment.
Regulatory-Ready Evaluation
Quality management, documentation, and reproducible benchmarks aligned with clinical practice.
Explainable Federated AI
Cross-institution learning with domain shift handling, robust aggregation, and governance-aware workflows.
Agentic AI Systems
Multi-agent orchestration for iterative analysis, conflict resolution, and transparent synthesis.
Multimodal LLM/VLM
Radiology report generation, VQA, and structured reasoning over images + text with retrieval.
People
Prof. Ghada Khoriba
Visiting Professor (Nile University)
Prof. Mohamed Mabrok
Visiting Professor (Qatar University)
Ahmed Elboardy
PhD Student (D2)
Yiming Jia
PhD Student (D1)
Mina Yousef
PhD Student (D1)
Makoto Kanehira
MSc Student (M2)
Ziad Elshaer
Research Student (RS)
Ka Ho Ng
Research Student (RS)
Joining soon...
Joining soon...
Selected Publications
📝 MaxVQA: A Large-Scale Multi-Source Dataset for Explanatory Visual Question Answering in Mammography (2026)
📝 Advances in Medical Image Segmentation: A Comprehensive Review with a Focus on Lumbar Spine Applications (2025)
Selected Awards
🏆 Student's Top Achievement Award
🏆 MEXT Science and Technology Award (Development Category)
Join us
We’re looking for
- We are seeking highly motivated and intellectually curious students with a strong foundation in mathematics, programming (e.g., Python), and machine learning, who are passionate about advancing AI-driven medical imaging research and are eager to engage in rigorous, collaborative, and publication-oriented research within an international environment.
How to apply
- Email your CV + short research statement
- Include links to papers / code (if available)
- Write 2–3 topics you want to work on
Tip: If you’re applying for scholarships (JSPS/MEXT/etc.), mention the program and deadline.
Contact
Lab Information
Affiliation: Graduate School of Information Science, University of Hyogo
Address: 7-1-28 Minatojima-minamimachi, Chuo-ku, Kobe, Hyogo 650-0047, Japan
Room: Lab. number 601 (6th floor)