How is OWKIN using machine learning to augment medical and biology research?
OWKIN Socrates is the first data-driven machine learning platform for medical research. It is designed to augment medical researchers’ skills and recapture the excitement of research and exploration.
Socrates will help researchers in academia, hospitals and the pharmaceutical industry. The platform is smart and will allow researchers to become machine teachers, without needing to understand the mathematics behind the scenes. The platform learns while the researchers discover, improving its global, collective intelligence.
The underlying machine learning technology is a unique integration of models built upon medical images, genomics and clinical data, allowing for the discovery of biomarkers and mechanisms associated with diseases and treatment outcomes. Our proprietary federated learning technology will enable collaboration among researchers and centres without sharing data or uploading it to the cloud.
Are medical professionals embracing such new tools in such a traditional industry?
The reception has been great. We’ve focused on making our product integrate into clinicians and researcher workflows. Machine learning is part of the next wave of technologies, and Socrates lets medical professionals be a part of that without needing to develop PhD computer science skills. We’ve definitely seen them embrace the tool, to the point where in its beta stage we’re having to be selective as to research questions.
How will the use of this data enhance patient care?
A) One of the key problems facing the pharmaceutical industry is that it isn’t able to fully leverage the data it generates. Another is that access to real-world data is expensive. OWKIN is solving these problems with collective intelligence built on real-world patient-generated data.
We partner with pharmaceutical researchers to use the platform’s knowledge to answer questions that will help cure diseases, without sharing any data from our hospital partners.
A typical use case with pharma would be utilising our technology - and the intelligence pre-trained on real world data - to better understand drug development opportunities using patient cohorts based on drug resistance and toxicities. We can help a company fine-tune patient recruitment in a clinical trial with imaging or other biomarkers. Pharma companies can also purchase predictive models built with real-world data, biomarkers or targets developed with our partners.
Who are some of the providers or organisation’s that OWKIN is working with, etc?
We are working with some of the largest pharmaceutical companies such as Amgen and Roche and have research collaborations with Institut Curie and INSERM (the second largest cancer research organisation in the world after the NCI), among many others.