Long underestimated, skin health is gaining ground in the era of image queen. According to the Xerfi firm, its turnover is expected to grow by 8.5% per year worldwide until 2027. A market of the future in which QuantifiCare wants to bring a differentiating approach. At the origin of the collaborative project PLICIA, acronym for Clinical Imaging Software Platform integrating Artificial Intelligence, the biotech based in Sophia Antipolis wants to become the world reference in the analysis of medical images.
Alone in the world (or almost)
Founded in 2001, the company is a spin-off from INRIA, a public research institute in computer science and automation. Its specialties: medical image processing, 3D systems and AI from which it develops services and products dedicated to diagnosis. It thus works with two types of clientele: pharmaceutical laboratories on the one hand, for which it evaluates, notably via scoring, candidate drugs and treatments during clinical studies, aesthetic dermatologists and plastic surgeons on the other, to whom it provides solutions allowing 3D simulations to be carried out before and after the intervention. A niche occupied by only two companies in the world, including QuantifiCare which deploys this dual activity in around sixty countries through 5 subsidiaries (2 in Brazil, one in the USA, the Philippines, Japan) and sales offices in Korea or still in Dubai. In total, the biotech employs 140 people and in 2019 raised a total budget of 10 million euros, including 5 million in equity from LBO France, to consolidate and strengthen its positions.
Innate Pharma’s strategy to carve out a place among immuno-oncology biotechs
Data anonymization
With PLICIA therefore, QuantifiCare wants to further explore the possibilities given by artificial intelligence. “ AI diagnosis is already on par with, and in some cases even surpassing, that of medical experts. By strengthening the tool in a very specific way and combining it with the perceptions of doctors, which remain essential, we can still make enormous progress, in particular towards subpopulations which, because they present this or that characteristic, escape or do not have as effective treatment as the rest of the population », Explains Jean-Philippe Thirion, the founding director of the biotech.
But to do this, an imperative is essential, that of the anonymization of the data from which the AIs learn. This is the whole purpose of the consortium of which Quantificare has taken the lead, made up of the SME Hosteur, specialized in data hosting, INRIA and the I3S laboratory (for Computer Science, Signals and Systems of Sophia Antipolis) which will provide its expertise in deep learning, machine learning and image processing. The group obtained financial support of 3 million euros over 3 years as part of the 2nd i-Démo call for projects, supported by the South Region and resulting from France 2030. “ The envelope remains modest but the partners have significant firepower “. Above all, the project differs from the approach of Gafam, whose “ security breaches have recently cost them some 3.4 billion euros in fines “, says the manager.
Innate Pharma’s strategy to carve out a place among immuno-oncology biotechs
Trusted third party
« PLICIA is built around 3 bricks, explains Jean-Philippe Thirion. The first focuses on the collection of very broad medical data from hospitals and doctors, in a secure manner. The second takes the form of a data lakea closed space where learning AIs are run. Nothing comes out except the data learned from the images. The doctor just needs to send an image taken with his phone, for example, and the system works to determine the pathology or give a pathological score. Impossible to trace to the source. Finally, the third allows the different AI to operate within the system.. » A productive and collaborative safe in short which aims to avoid widening the gap vis-à-vis American and Chinese approaches to AI while respecting European ethics. The consortium also intends to open the system to university hospitals with a lot of medical data without knowing what to do with it or how to process it. “ It is a secure data storage tool that will help advance science and diagnostics “. In other words, a trusted third party whose first PoCs are in progress. The final, deployable version is expected in 2027.
Related News :