cv

Education, experience, and skills — download the full PDF via the icon above.

General Information

Full Name Romain Andres
Location Caen, Normandy, France
Languages French (native), English (B2, CLES B1 certification)

Education

  • Master's degree in AI and Data Analysis, focus on Healthcare

    2023 - 2025

    University of Caen
    • Part of the SATIN project (Health, Territory, Innovation and Digital Technology).
  • Bachelor's degree in Computer Science

    2023

    University of Caen
  • First Aid Certificate (PSC1 — Prevention and Civic Rescue, Level 1)

    02/2024

    University of Caen
    • First aid training providing the skills required to respond in the event of a medical emergency.
  • French Baccalauréat, Economics and Social Sciences stream

    2019

    Lycée Fresnel, Caen

Experience

  • AI Engineer — Federated Learning for Medical Imaging

    09/25 – 06/26

    Centre François Baclesse, Caen
    • Fixed-term contract contributing to a national federated learning project advancing personalized medicine through improved survival prediction for glioblastoma patients, building on prior Master's work in explainable AI.
    • Collaborated with a multi-institutional team from Inria and several French cancer centers to design and implement technical solutions.
    • Trained a deep learning model on distributed data using the Fed-BioMed framework, ensuring patient data privacy throughout.
    • Managed the project's data pipeline, including standardization and anonymization of multi-center datasets.
    • Addressed complex architectural challenges inherent to federated learning environments.
  • Research and Development Intern, Explainability

    03/25 – 08/25

    Centre François Baclesse, Caen
    • Explored and applied explainability techniques (Grad-CAM, LIME, LRP) to understand the predictions of a ResNet-51q model for glioblastoma treatment response.
    • Ran statistical analysis and validation against clinical data to ensure the relevance and robustness of the generated explanations.
    • Compared explainability methods and identified the model's limitations across patient cases.
    • Wrote and submitted a scientific article presenting the research outcomes to a peer-reviewed venue.
    • Skills strengthened
      • Deep learning & explainability (ResNet, Grad-CAM, LIME, LRP)
      • Medical image analysis (MRI, segmentation, feature extraction)
      • Statistics and validation of AI models applied to healthcare
      • Scientific writing and the publication process
  • Research and Development Intern, Medical AI

    05/24 – 07/24

    Centre François Baclesse, Caen
    • Continuation of a Master's project creating a visualization interface and API for a deep learning algorithm for automatic brain lesion segmentation on MRI, under the supervision of Dr. Corroyer-Dulmont (Head of AI and Medical Imaging Engineer at CFB).
    • Deployed, finalized, and optimized an interface based on OHIF Viewer and its API on the center's Windows environments (ML Ops).
    • Optimized and retrained the deep learning model on a new dataset to reduce computational cost.
    • Outcome
      • Co-first author, "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).
    paper
  • Front-end Web Developer

    06/23 – 07/23

    Atlas Music, Assas Lab Incubator, Paris
    • Implemented Cypress E2E tests and front-end development with React, GraphQL, and MongoDB, with a focus on responsive design.

Publications

  • Development and routine implementation of a deep learning algorithm for automatic brain metastases segmentation on MRI for RANO-BM criteria follow-up

    2025

    NeuroImage, Volume 306 (2025) — co-first author
    paper
  • Explainability and Clinical Trust of Deep Learning in Glioblastoma Treatment Efficacy Prediction: A Comprehensive Framework

    2025

    Manuscript under review — first author
    code

Skills

  • Deep Learning & Computer Vision
    • PyTorch, TensorBoard
    • Convolutional neural networks (ResNet, UNet, UNet-R)
    • Explainable AI — Grad-CAM, LIME, LRP, linear probing, causal activation patching
  • Medical Imaging
    • DICOM, NIfTI, HDF5 (H5) formats
    • Image preprocessing, segmentation, feature extraction on MRI / CT-scan data
    • Federated learning for medical data (Fed-BioMed)
  • Data Analysis & Machine Learning
    • Python: NumPy, SciPy, scikit-learn, pandas — plus R
    • Statistical testing, model evaluation, robustness analysis
    • Supervised, unsupervised and reinforcement learning, SVMs, dimensionality reduction
    • Data mining and ETL pipelines (KNIME, Weka, Talend)
  • Databases
    • PostgreSQL, MongoDB, Cassandra, MySQL, MariaDB
    • GraphQL
  • Development & Systems
    • Git, Linux, Bash, Docker, Kubernetes
    • React / Flask for research prototyping and medical imaging interfaces
    • Self-hosted FreeBSD NAS/cloud server administration

Additional Knowledge

  • Research methodology — literature review, academic writing (LaTeX), reference management (Zotero), reproducible experimentation, peer-reviewed publication process.
  • Scientific communication — peer-reviewed publication (co-first author, NeuroImage), poster presentations, public speaking, conference hosting.
  • Biomedical knowledge — telemedicine, physiology, neurology, genetics, ECG analysis, pharmacology (SMILES/SMARTS notation, molecular graph representations).