Portfolio & CV · Publications · LinkedIn · Email
I build deep learning systems for oncology that clinicians can actually trust — from segmentation models deployed in radiotherapy workflows to explainability pipelines that catch a model relying on the wrong signal before it reaches a clinic. Co-first author on a NeuroImage paper, first author on a manuscript currently under review. Currently looking for my next step: a CIFRE PhD or a research/AI engineer role.
 | Explainability & Clinical Trust of Deep Learning in Glioblastoma Treatment Response — first-author manuscript, under review A multi-layered XAI pipeline (Grad-CAM, LRP, LIME, linear probing, causal activation patching) auditing a ResNet-51q model — and catching it relying on a proxy for surgical resection status instead of real tumoral features. |
 | MetIA — Deep Learning Interface for Brain Metastases Segmentation — co-first author, published in NeuroImage, Vol. 306 (2025) UNETR-based segmentation model deployed into a clinical OHIF Viewer interface at Centre François Baclesse, from Flask API to ML Ops on the center's infrastructure. |
 | GeneticPedigreeChartToPed — Digitizing Family Trees for Hereditary Cancer Risk A six-model computer vision pipeline (YOLO, EasyOCR, DeepLSD, graph reconstruction) turning hand-drawn pedigree charts into structured data for tools like CanRisk. |
🗂️ Earlier projects (coursework & side projects)
| Project | Description |
|---|---|
| Sorting Algorithms Visualizer | C++ visualizer + Python/Jupyter benchmarking of sorting algorithms across data distributions. |
| AI vs. AI — Virus Board Game | Minimax/Alpha-Beta Pruning agents battling on a custom board game. |
| Todolist — React Native | First React project: a to-do app on a Node.js/GraphQL CRUD API. |
| Fractal Flowers Generator | Procedural flora generation from scratch using L-systems (Java). |
| Title | Venue | Role |
|---|---|---|
| Development and routine implementation of a deep learning algorithm for automatic brain metastases segmentation on MRI for RANO-BM criteria follow-up | NeuroImage, Vol. 306 (2025) | Co-first author |
| Explainability and Clinical Trust of Deep Learning in Glioblastoma Treatment Efficacy Prediction | Manuscript under review (2025) | First author |
Full abstracts and BibTeX on the publications page.
- Deep learning & XAI: PyTorch · Grad-CAM/LRP/LIME · UNet/UNETR/ResNet · Fed-BioMed
- Medical imaging: DICOM · NIfTI · OHIF Viewer
- Data & backend: Python (NumPy/SciPy/pandas) · Flask · GraphQL · PostgreSQL · MongoDB
- Tools: Docker · Kubernetes · Git · Linux
