Jonas Mehtali
PhD Student in Computer-Assisted Interventions · ICube Laboratory, University of Strasbourg
PhD Student — ICube Laboratory
Automatic multi-needle adaptive planning for percutaneous thermal ablation.
PhD since September 2024 · Caroline Essert and Juan Verde · Strasbourg, France
Research
Thermal ablation is an established option for treating small liver tumors. Comparable oncologic outcomes to surgery, minimally invasive, low morbidity. The problem: when a tumor exceeds what a single needle can cover, clinicians plan multiple overlapping ablations using 2D CT views and complex interfaces. Local recurrence rates sit between 10 and 39.1% within five years, often because adequate coverage is hard to achieve manually.
My PhD addresses this across three axes, built directly on my Master internships at ICube and IHU Strasbourg. Currently funded through the ITI HealthTech doctoral program at ED269 MSII.
Multi-needle trajectory optimization
Compare exhaustive and learning-based methods for optimal needle placement, accounting for vital anatomical structures. Explore search space reduction, adaptive simulation resolution, and DL-based ablation volume estimation.
Intraoperative replanning
Extend optimization to real-time replanning. Track already-inserted needles, estimate partly treated volume dynamically, and adapt the plan for subsequent insertions as the procedure unfolds.
Ergonomic planning interface
Design and implement a user-friendly GUI for planning, tracking, and replanning. Validate usability through studies with both clinical experts and novices.
Technical Toolkit
| Category | Tools |
|---|---|
| Languages | Python · C/C++ · CUDA |
| Medical Imaging | 3D Slicer · PyTorch · VTK |
| Graphics & XR | OpenGL · Unity · Blender · Maya |
Publications
| Paper | Venue | Highlights | DOI |
|---|---|---|---|
| Cryotrack — Planning and Navigation for Computer Assisted Cryoablation | MICCAI 2024 | H.J. Krumb, J. Mehtali, J. Verde, A. Mukhopadhyay, C. Essert | 10.1007/978-3-031-72089-5_10 |
| HEAT — High-Efficiency Simulation for Thermal Ablation Therapy | IJCARS 2025 | J. Mehtali, J. Verde, C. Essert — GPU-accelerated multi-resolution simulation, results in under 1 second | 10.1007/s11548-025-03350-z |
| C-NCA — Chained Neural Cellular Automata for Fast Thermal Ablation Estimation | MICCAI 2025 Latest | J. Mehtali, J. Verde, C. Essert — 12,210 params · 476 fps · interactive optimization loops | 10.1007/978-3-032-04965-0_7 |
Beyond Research
When something annoying needs solving, I build tools for it. Small utilities, experiments with LLMs, whatever makes a recurring problem less tedious. Side projects include game development: started with Sensha (Python/PyGame) for my Baccalauréat, later revisited in Godot with multiplayer support.
Profiles
- GitHub — JonasMht
- Google Scholar — Jonas Mehtali
- ORCID — 0009-0003-3688-7782
- ResearchGate — Jonas Mehtali
- HAL — Jonas Mehtali