Looking for a job

Claudiu Ivan

@claudiuivan

Message

I build high-performance machine learning software for medical image analysis and 4D fMRI research.

Germany

Message

I've built high-performance scientific software for medical image analysis at the Institute of Experimental and Clinical Pharmacology and Toxicology at FAU Erlangen-Nürnberg since 2015. My work has reduced ReHo computation from days to seconds using FFT-based convolutions and cut a Kamada-Kawai graph-layout workflow from days to seconds through convex optimisation.

I lead low-level deep learning work for 4D fMRI analysis, designing and evaluating CNN, RNN, and hybrid architectures in Python with TensorFlow and PyTorch. I run large-scale experiments on university HPC infrastructure, use Weights & Biases for experiment tracking and hyperparameter management, and help scale ANTs image-registration workflows.

I've presented research at Neuroinformatics 2016 and EMIM 2021, and I support colleagues with programming, machine learning methodology, and technical troubleshooting.

Experience

Work history, roles, and key accomplishments

Medical Researcher

Institut für Experimentelle und Klinische Pharmakologie und Toxikologi

Apr 2015 - Dec 2024 (9 years 8 months)

Built high-performance scientific software for medical image analysis, including custom IDL routines for NIfTI and proprietary scanner formats. Implemented algorithms from research literature and delivered major performance gains: reduced ReHo computation from days to seconds with FFT-based convolutions, and reduced a Kamada-Kawai graph-layout workflow from days to seconds

research assistant

Institute of Experimental and Clinical Pharmacology and Toxicology

Apr 2015 - Dec 2024 (9 years 8 months)

Conducts research in experimental and clinical pharmacology and toxicology, applying machine learning and medical image analysis techniques.

Intern

Institut für Experimentelle und Klinische Pharmakologie und Toxikologi

Jan 2013 - Jan 2013 (0 months)

Competitively selected for a 3-month Erasmus placement between 3rd and 4th year of B.Sc. Self-taught IDL from scratch and delivered medical image processing software that outperformed tools built by practitioners with 10+ years of experience — earning an invitation to return for the M.Sc. and join the group long-term.

Data Preparation

Actor

Freelance

Jan 2007 - Jan 2010 (3 years)

Performing arts training at Kunstschule Brasov (2005–2009); award-winning work at national theatre festivals. Developed strong communication, presence, and teamwork skills.