About Me

👋 Short Bio

Hi, I’m Ziad Kheil.

I recently completed my PhD (December 🎉) at IUCT-Oncopole and CRCT, where I worked on deep learning for oncology. My PhD centered on learning-based deformable image registration and physics-informed modeling.

Previously, I worked on robust deep learning at IRT Saint Exupéry. I hold an MSc in Applied Mathematics & Computer Science from ISAE-Supaero and an MSc in AI & Deep Learning from Imperial College London.

I care about methods that translate to real clinical workflows.

🔬 Research Focus

My research explores knowledge-guided, constraint-aware deep learning, motivated in part by challenges in radiotherapy planning (motion, alignment, physics-aware models). This has naturally led me from deformable registration to multimodal learning across imaging, whole-slide pathology, and transcriptomics.

More broadly, I’m interested in models that support clinical decision-making and yield actionable biological insight from multi-modal data.

💡 Themes (non-exhaustive)

• 🫁 Deformable image registration
• 🏆 Foundation models in medicine
• ⚠️ Clinical robustness
• 🧲 Physics-informed neural networks
• 🎨 Generative modeling
• 🧬 Beyond imaging: multimodal learning

Generally speaking I value clinical usefulness over leaderboard performance.

Outside the lab

• ♟️ Chess — enthusiastic improver, still trying to navigate the Pirc.
• 📚 Reading — philosophy, science, and anything that rewards curiosity.
• ☕️ Coffee — the most consistent part of my training pipeline.

Feel free to reach out — I’m always happy to grab a coffee ☕️