AI subgroup discovery
Multimodal AI-powered biomarkers
We combine cutting-edge machine learning and biology to identify biomarkers
Challenge
Precision medicine demands a deeper understanding of patient biology
- Today's biomarkers are unimodal and limited to one aspect of the patient information
Solution
Multimodal AI-powered biomarkers
- Designed to answer the most pressing research questions and to power our AI engines
Owkin AI-powered biomarkers power precision medicine
Owkin multimodal AI-powered biomarkers
Discovery
Clinical development
Clinical routine
Owkin’s solution
Our AI powered
orthogonal approach
Our orthogonal approach combines data from multiple modalities to fully understand the tumor microenvironment to define phenotypic subtypes.
Our biomarker models are built on high-quality patient data, carefully selected and verified by medical experts from our Federated Research Network.
These models have been designed to be interpretable by researchers and medical experts leading to a multiscale understanding of the disease.
Biomarker case study
Inclusion criteria models
Owkin’s inclusion criteria models increase the probability of phase II/III trial success through better defined patient populations
Methodology
Click
1
drug mechanism of action
Identify biomarker required to define subgroups of patients.
Step 1
Understand
drug mechanism of action
Click
2
inclusion criteria model
Either choose existing Owkin model to select patients or build bespoke model from multimodal patient data to identify the right patient population.
Step 2
Select/build
inclusion criteria model to select patients
Click
3
model to digitized H&E slides of trial participants
Select specific patient population for trial recruitment most likely to respond to treatment.
Step 3
Apply model
to digitized H&E slides to select participants