Alexandre Filiot

Alexandre Filiot

Computer vision @ Owkin

CV

My CV is available here.

Work experience

Senior data scientist @OWKIN, Sep. 2022 - now

I am part of the Medical Imaging group which focuses on the analysis and representation of histopathology data. I work within R&D on self-supervised learning to better represent digital pathology images, leveraging state-of-the-art computer vision and ML techniques to fuel all our internal research projects and pathology AI solutions. I am one of the main contributor behind Phikon, Phikon-v2, H0-Mini and other internal models. More specically, I am also involved in the development and continuous improvement of our histology-based, AI-guided diagnostics tools such as RlapsRisk BC (lately MSIntuit CRC). Also,

Data scientist, Nov. 2019, Aug. 2022 - Lille University Hospital | INSERM | INRIA

Data scientist @ENDOMIC, Dec. 2021 - Aug. 2022

Statistical learning and image analysis applied to clinical research as part of the ENDOMIC team (INSERM, INRIA, Lille University). Working on

  • Digital histopathology: whole slide image analysis using deep learning techniques applied to esogastric cancers (molecular subtyping, outcome prediction, etc.)
  • Immunofluoresence imaging: weekly supervised learning and cluster analysis applied to IIF images on HEp2-cells for autoimmune diseases (patterns recognition, auto-antibodies classification, etc.)

Data scientist @INCLUDE, Nov. 2019 - Nov. 2021

Statistical-learning-based clinical research projects as part of the INCLUDE team (INtegration Center of the Lille University hospital for Data Exploration). I provided statistical support for clinical research projects (biostatistics, machine learning), covering data mining and protocol design, implementation, results sharing and feedbacks with clinicians, publication if relevant. Also working on long-term projects:

  • Digital histopathology: whole slide image analysis using deep learning techniques applied to esogastric cancers (molecular subtyping, outcome prediction, etc.)
  • Federated learning: POCs on the applicability of federated learning to medical research in close collaboration with the Inria Magnet team, IT engineers, DPOs, etc. Winner of the 2021 call for proposals Bac-à-sable CNIL.

Research intern @Guerbet, Jun. 2019 - Nov. 2019

  • GANs: virtual contrast enhancement applied on brain MRIs using deep generative networks.
  • 3DCNN: gliomas segmentation on Brats data set using 3D convolution networks.

Data scientist intern @Banque de France, Jun. 2018 - Sep. 2018

  • Credit scoring: statistical modelling and 3 year-prediction of the European Fractional Default Risk.
  • Natural Language Processing: analysis of companies’ financial reports to evaluate the semantic risk of credit default.

Education