2008 - 2012Working as a R&D engineer and acted as the bridge between engineering and research teams. I was involved in research and engineering projects in the medical imaging field, mainly in the oncology and cardiology area. The tasks carried out included:
• Research and development of algorithm prototypes in matlab and its technical documentation. I developed efficient algorithms using machine learning techniques (SVM, SURF) to detect points of interest in CT images. I also worked on prototypes of interactive deformable registration to register CT/CT images,
• Efficient implementation of algorithms from technical documentation in C++. Some modules involved low level optimization (SSE2 for low level image processing) and multithreading (OpenMP in a Region of interest detector),
• Design, implementation and testing of complex engineering components in C++/C#/Java,
• Requirement, design and test specification documents writing,
• Tooling improvements. As an example, I developed a matlab-like tool with its own programming language (parsing and interpreting a subset of C++) to rapidly develop medical imaging prototypes (using C++, STL and Boost.Thread).
• Ramp-up of new team members.
I successfully took part to all the stages of the product lifecycle and improved my knowledge in medical imaging and image processing. I participated to projects (500+ developers) involving cross-site activities (USA, India, and Germany) and agile development. My work followed high quality processes in a controlled development environment, necessary to enforce FDA regulations.
2007 - 2008One year internship in the Imaging and Visualization department. I worked in the area of medical imaging and implemented state-of-the-art algorithms such as smoothing surface and mesh simplification involving massive data.
I also worked alone (8 months) on a research project: Automatic skeleton partitioning and recognition. This project involved the use of cutting edge technologies like Support Vector Machines, neural networks, hidden markov models, genetic algorithms, simulated annealing, clustering and more generally machine learning and image processing techniques. I have improved my skills in linear algebra and statistical data modeling. This generic framework had been developed in C++/STL using Visual Studio 2005.
I managed to achieve good results under loose supervision where I could try any ideas. The algorithms are patented and used in several products. My work included:
• Research, design and implementation of effective algorithms.
• Prototyping demonstrations.
• Writing user documentation, technical paper and invention disclosure.
EPITA
- Assistant teacher
Villejuif2006 - 2007Teaching programing languages to 300 third year students at EPITA (computer science engineering school, 5 year degree)
