Francois Chadebecq

Post Doc.


I obtained a Master’s Degree in Models, Systems, Imaging and Robotics from University Blaise Pascal of Clermont-Ferrand in 2010 with a major in Computer Vision. I then realised a PhD, which I obtained in 2015, under the supervision of Pr A. Bartoli and Dr C. Tilmant with a dual affiliation to the ComSee group at Institut Pascal and the ALCoV group at Image Science for Interventional Techniques laboratory in Clermont-Ferrand. I am currently a research associate involved in the fetal surgery project (GIFT-Surg) funded by EPSRC and Wellcome Trust. My main research interests focus on the development of innovative computer Vision techniques for medical imaging and more particularly endoscopic surgery. My PhD reseach aimed at developing a passive computer vision method to measure the size of tumours in colonoscopy. With the help of my supervisors, I proposed an innovative method adapted to the prime lens optical systems of colonoscopes, to extract the sharp/blur breakpoint (named the Infocus-Breakpoint) in a video stream obtained by moving the colonoscope toward a neoplasia (Chadebecq et al., MIA 2014). I did several experiments in collaboration with endoscopists (Hospices Civils de Lyon) to validate the proposed approach in real colonoscopic conditions. I furthermore designed a phantom model and conducted synthetic evaluations which showed that the measurement error is less than a millimetre which represents a percentage of error lower than 7% (Chadebecq et al., ISBI 2013). This significantly outperforms the visual estimation method currently used by endoscopists for which (Chaptini et al., Journal of Gastrointestinal Endoscopy 2014) showed that for more than half of the cases the estimation error is greater than 20%.


  • 2017

  • science!

    Adjoint Transformation Algorithm for Hand-eye Calibration with Applications in Robotic Assisted Surgery (currently under revision)

  • 2015

  • science!

    Practical Dry Calibration With Medium Adaptation For Fluid-Immersed Endoscopy