HomeTechnologyArtificial intelligenceNature Medicine publishes breakthrough Owkin

Nature Medicine publishes breakthrough Owkin

image: Jean du Terrail, Senior Machine Learning Scientist at Owlin, and lead author of new research published today in Nature Medicine
display Lake

Credit: Okkin

In research published in Naturopathy TodayAI biotech company Owkin has demonstrated for the first time that federated learning (FL) can be used to train deep learning models on multi-hospital data on histopathology data without the data leaving hospital firewalls.

The discovery paves the way for AI-powered medical research using larger multicentric data sets, allowing models to escape the biases of single-centric studies. This has the potential to unlock breakthroughs in precision medicine through the use of secure and privacy-protecting AI.

Using data held at four leading French hospitals, Owkin has built AI models that can accurately predict the future response of patients with triple-negative breast cancer (TNBC) to neoadjuvant chemotherapy. By using interpretable AI to extract information from digital pathology slides, Owlin was able to find potential new biomarkers. In the future, this could help guide patients to less toxic treatments or to new experimental treatments, allowing better tailor-made medical care.

The project used federated learning – a collaborative AI framework that preserves data privacy and security through Substra, Owkin’s recently open source software that makes every operation traceable using hyperledger technology. The study marks the first time machine learning models have been trained using histopathology data from multiple hospitals without the data leaving the hospitals. Previously, most studies were limited to simulating FL by artificially splitting data. It is a groundbreaking proof of FL in medical research and represents a breakthrough in realizing the practical benefit of AI for research.

The study used digital pathology data and clinical information from 650 patients from Institut Curie in Paris, Center Léon Bérard in Lyon, Gustave Roussy in Villejuif and IUCT Oncopole in Toulouse, making it one of the largest TNBC cohorts of its kind ever conducted for a such kind are composed. of analysis.

The research builds on Owlin’s pioneering use of FL to enable pharmaceutical companies to collaborate on drug discovery research while ensuring privacy, security and competitive considerations. The results of the MELLODDY project, published this year, showed that AI collaboration for drug discovery at an industrial scale is possible thanks to FL, an industry first. In addition to addressing privacy and security concerns, FL can also simplify data management, eliminate the need to transfer data, and encourage more collaborative research.

Jean du Terrail, Lead Author and Senior Machine Learning Scientist at Owlin, said:

Thanks to our partners, we are proud to have performed an original federated analysis of real-world medical data, and the first of its kind on histopathology data. By connecting institutions in a federated way, we were able to reach the critical mass of triple-negative breast cancer data that the AI ​​needs to discover on its own histological patterns predictive of treatment response. We hope this proof of concept will inspire medical institutions to collaborate in federated learning networks to advance research while keeping patient data private.

Julien Guérin, Chief Data Officer at Institut Curie in Paris, France, said:

We have reached an important milestone with the deployment of this federated learning infrastructure, which demonstrates a new advanced approach to building AI in cancer research. We are very happy to have been part of this adventure and hope that it will open promising perspectives for the future of patient care.

Dr. Guillaume Bataillon, pathologist at IUCT Oncopole in Toulouse, France, and former pathologist at Institut Curie in Paris, France, said:

Through this multidisciplinary partnership, we verify the feasibility of a collaborative federated learning approach between hospitals for a relevant biological question. This allowed us to create a pooled heterogeneous dataset in a safe and faster way to develop reproducible, transferable and even interpretable models. This proof of concept has the potential to become an aid to therapeutic decisions.

Dr. Pierre Etienne Heudel, medical oncologist at Center Léon Bérard in Lyon, France, said:

The emergence of digital pathology coupled with the explosion of various machine learning techniques should enable ever more accurate and personalized medicine. In addition, the federated learning achieved in this project by avoiding external data streams facilitates and secures the process for future day-to-day clinical practice.

Dr. Magali Lacroix-Triki, pathologist with Gustave Roussy in Villejuif, France, said:

Digital pathology and AI represent the third revolution in the world of pathology, and pathologists are excited to lead this new change in their practice. Federated learning, pioneering AI research in digital pathology, takes us one step closer to identifying novel biomarkers in oncology while ensuring data privacy and security.

Dr. Camille Franchet, pathologist at IUCT Oncopole in Toulouse, France, said:

By enabling AI models to be trained on multicentric data without centralization, federated learning unlocks one of the biggest obstacles in machine learning on medical data without compromising respect for personal data.

About Okin

owl is an AI biotechnology company that uses artificial intelligence to find the right treatment for every patient. We bridge shared innovation challenges between biopharmaceutical and academic researchers and close the translational gap between complex biology and new drugs.

We use AI to identify new treatments, reduce risk and accelerate clinical trials and build diagnostic tools that improve patient outcomes. Using federated learning, a breakthrough collaborative AI framework, Owkin enables medical and biopharmaceutical partners to unlock valuable insights from isolated datasets while protecting patient privacy and securing proprietary data.

Owkin was co-founded in 2016 by Thomas Clozel MD, a former assistant professor of clinical onco-hematology, and Gilles Wainrib, a machine learning pioneer in biology. Owkin has raised over $300 million and became a unicorn through investments from leading biopharmaceuticals (Sanofi and BMS) and venture capital funds (Fidelity, GV and BPI, among others).

disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of press releases posted on EurekAlert! by contributing institutions or for the use of information through the EurekAlert system.

Must Read
The Global Call Center AI Market is expected to grow by $ 1.90 bn during 2021-2025, progressing at a CAGR of 22.40% during the forecast period


Please enter your comment!
Please enter your name here

Most Popular

Recent Comments