Drive medical research with machine learning insights
Owkin Studio is a software application dedicated to medical researchers who want to use machine learning on healthcare data to discover new biomarkers, build prognosis models, predict response to treatment, and understand the results from a biological perspective.
Easily build machine learning projects with Owkin Studio.
A ground breaking approach to do research
Autonomously set up machine learning projects in a snap
Build cohorts and specify medical hypotheses. Train algorithms developed by our world class data scientists. You don’t have to be an expert yourself to use Studio - no coding required.
Uncover how your model works
Understand how a Machine Learning model analyzed your data thanks to Studio’s interpretability features. Integrate your medical knowledge to draw your own conclusions and refine your research.
Take your project to the next level
Studio is a collaborative platform where researchers and data scientists come together. Advance your research further by integrating custom models that are tailor made to your research question.
How Owkin Studio works
Explore how to manage your Al-powered research project
New biomarkers in MM through Owkin’s machine learning
Malignant mesothelioma (MM) is an aggressive cancer. The tumor can be classified into three histological types: epithelioid, biphasic and sarcomatoid MM, each of them associated with an average prediction of the patient outcome.
In 2018 and 2019, Owkin collaborated with Francoise Galateau-Sallé, Department of Biopathology at Cancer Center Léon Bérard in Lyon, to identify new histological features predictive of the survival of the patients.
MesoNet pointed out regions located in the stroma, that were associated with low survival.
MesoNet is a good example of how machine learning can provide tools and visualization to identify new biomarkers!
Pierre Courtiol – Data scientist
Analysis of MesoBANK Images and clinical data: 2300 whole slide images, along with patient's tumor grade status and histological subtype…
The algorithm evaluated all the tiles of the dataset to identify the ones predictive of survival
Our pathologist partner at CLB went through single tiles and identified new morphological markers associated to prognosis
Explore how the model works by analyzing the tiles that display a low and high survival score. Can you find the morphological markers that allow you to precisely predict patient survival?