To progress things to the next level, precision medicine, or personalized medicine, must make use of artificial intelligence. Pathologists are gradually beginning to accept this change, albeit without fully mastering it. The PortrAIt Consortium will help them with this tricky transition. “This gradual transition to digital technology is recent,” said Meriem Sefta, chief diagnostics officer at Owkin. “The deployment of AI tools raises many questions. These professionals require support and training.” This support has helped Dr Magali Lacroix-Triki, pathologist at Gustave Roussy, a stakeholder in the Consortium. “I didn’t know anything about AI, but have learned a lot as a result of this collaboration with Owkin,” said the breast cancer specialist, who is president of the French Division of the International Academy of Pathology.
Created in July 2022, the project was officially unveiled in late March 2023. Over the next five years, PortrAIt aims to create at least 15 AI tools for digital pathology, which may speed up pathologists’ work, detect the presence of certain biomarkers, and predict patient outcomes.
There is strength in numbers and in order to meet this challenge, Owkin, the global specialist in AI applied to medicine, has surrounded itself with the best oncology and data science experts – Gustave Roussy (the largest cancer center in Europe), Unicancer, the National Federation of Cancer Centers (whose centers will contribute to the work, in particular via their pathology departments), the Léon Bérard Center (cancer treatment center in Lyon), Tribun Health, the French digital pathology workflow expert and Cypath (an independent pathology group).
“Digital pathology-based precision medicine promises to transform the way we diagnose and treat diseases. It is enabling doctors to tailor treatment to the specific needs of their patients at an earlier stage while democratizing access to high-quality, personalized treatment. PortrAIt will empower French anatomical pathology laboratories to play a leading role in the development of the next generation of AI-powered diagnostic tools. Our aim is to make France the global leader in digital pathology-based precision medicine,” Meriem Sefta, chief diagnostics officer at Owkin.
PortrAIt will be broken down in two phases. Initially, a platform, PortrAIt Lab, will be created as early as July 2023. “Pathologists will have a role to play in developing diagnostic tools by annotating slides, for example, to train AI models,” said Etienne Hatton, Consortium project manager. “This large amount of data will also transform this lab into a support base for research. ”
Then, a second distribution platform will function a bit like a market place. This platform will be managed by Tribun Health and Owkin, and will act as a reliable resource for all operators involved in the sector, with the aim of creating a digital pathology channel. “The idea is to facilitate the availability of artificial intelligence tools for pathologists in France and, over time, abroad,” Etienne Hatton continued. This platform will provide fifteen or so AI diagnostic tools that can be routinely used when the project comes to an end in five years’ time. The platform is also designed to host tools from other AI enterprises. “We wish to offer pathologists the most appropriate and complete solution possible,” Etienne Hatton said in conclusion. Medical professionals will be able to make use of these tools on a daily basis in order to refine their diagnosis and save time, particularly in regard to cancer treatment. “Embedded algorithms will be present very early in the scanned slides visualization process, notably via the use of Tribun Health software,” said Dr Magali Lacroix-Triki. “Depending on the patient, pathologists may use these algorithms to obtain additional predictive information. ”
An initial collaboration between Owkin and Gustave Roussy has already proven its worth. In 2019, the startup won the “AI for Health” challenge organized by the Île de France Region, in partnership with Gustave Roussy. The research this initiated led to the launch, in September 2022, of RlapsRisk BC, an AI-based prognostic solution to predict the probability of relapse after treatment for women with early stage breast cancer (ER+/HER2-). This tool therefore allows oncologists to determine which high-risk patients may benefit from targeted therapies and which low-risk patients may potentially avoid chemotherapy. This has been “a fruitful and friendly collaboration,” according to Dr Lacroix-Triki, which has led to several RlapsRisk BC development and validation projects for breast cancer patients.
PortrAIt has a budget of €33 million, funded by France 2030 (operated on behalf of the state by Bpifrance) and by the European Union (Next Generation EU as part of the “France Relance” recovery plan).
This consortium was created in response to an I-Demo call for projects in connection with the France 2030 plan, with the aim of developing industrial and service businesses in growth markets that create value and competitiveness for the French economy, in this case the digital health sector.
In addition to breast cancer, projects will also look into biomarkers for other cancers, such as lung cancer.