Owkin, founded by Thomas Clozel (MD and oncologist) and Gilles Wainrib (PhD in machine learning applied to biomedicine), is an AI startup that builds machine learning tools to augment medical and biology research. Their proprietary platform Owkin Studio integrates biomedical images, genomics and clinical data to discover biomarkers associated with treatment outcomes, building models that are as interpretable as possible, for a deeper understanding of the initial data. Through transfer learning, Owkin push the intelligence of pre-trained algorithms into new models which will have a better predictive power from cohorts as small as 100 patients. Owkin’s international and multidisciplinary team, with exceptional talent in machine learning, data science, biology, medicine, operations and business is among the best in the world, with several Kaggle Masters, and a DREAM Challenge top performer. Owkin partners with top-tier oncology institutions in Europe (INSERM, Institut Curie, Centre Léon Bérard) and in the US (Mount Sinai), and leading pharmaceutical companies (Roche, Amgen, Ipsen), leading to high- ranking publications in top scientific journals (ICML, NIPS). Owkin is leading the Health Chain research consortium that was awarded a EUR 10M French grant for responsible and accountable AI blockchain.
In the H2020 MASTRIA project, Owkin brings its experience in multimodal data integration for relating radiomic features to occurrence of AF outcome and to build a radiomic signature of FA., acquired from multiple projects in diseases such as CVD events prediction, glioblastoma, or hepatocellular carcinoma, studying patient response to treatment or patient survival (overall, progression free, disease free). Owkin also has developed specific expertise in machine learning for radiology images images, with multiple deep learning algorithms able to work at different radiologic modalities and automated annotations, while remaining interpretable for radiologist and clinicians.