About Me

👋 Short Bio

Hi, I’m Ziad Kheil, a machine learning researcher working on deep learning for biomedical and physical data.

I did my PhD (December 2025) at IUCT-Oncopole and CRCT, building physical constraints into neural networks for radiotherapy, on clinical CT and MRI with medical physicists. I also worked on multimodal learning across MRI, pathology and omics. Before that: robust deep learning at IRT Saint Exupéry, ISAE-SUPAERO and Imperial College London.

Now: looking for ML roles in France, in biomedical AI and physics-informed ML. Get in touch.

🔬 Research Focus

I’m interested in models that build known structure in (physical laws, anatomy, biological priors) instead of hoping a network learns it from data alone, especially where data is scarce and mistakes are costly.

• Data I work with
  • 🫁 CT, 4D-CT and MRI
  • 🔬 Whole-slide pathology
  • 🧬 Transcriptomics and DNA methylation
  • 🧲 Physical systems
• Methods
  • 📐 Physics-informed and constrained learning
  • 🔗 Multimodal representation learning
  • 🏆 Foundation models
  • ⚠️ Robustness and failure analysis
  • 🎨 Generative models

Code and papers: see publications.

Outside the lab

♟️ Chess (still trying to navigate the Pirc) · 📚 philosophy and science · ☕️ coffee, the most consistent part of my training pipeline. Always happy to grab one.