In early April, Anap (the French national agency for supporting the performance of healthcare and medico-social institutions) announced the launch of a national platform to share artificial intelligence solutions deployed in healthcare institutions. It is possible to consult, for each of the referenced solutions, its history, its level of maturity, the number of users who benefit from it and the key success factors identified. The agency invites healthcare institutions to share their AI projects.

Anap, a public agency of expertise attached to the Ministry of Health, has the mission to respond to the needs of health and medico-social institutions through actions (methods, tools, events, interventions) developed with and for professionals.

Healthcare professionals use AI solutions (to help them diagnose and choose treatments, predict patient flows or automate tasks) to improve care, patient experience and the internal organization of facilities.

To support the deployment of these solutions, Anap is launching a national platform for sharing AI solutions, which will enable the referencing of projects carried out in healthcare institutions and the sharing of key information on their design and deployment, in order to inspire all healthcare stakeholders.

Stéphane Pardoux, CEO of Anap states:

“This platform is a sharing space open to all professionals. Our goal is to give visibility to the artificial intelligence solutions deployed and to capitalize on these experiences! Artificial intelligence is a major performance lever for healthcare institutions: to provide better care, to improve the flow of care and to save time by automating certain administrative tasks. I call on professionals to take advantage of our platform, to share their solutions or to come and find useful information
I call on professionals to take advantage of our platform, to share their solutions or to come and find useful information for their projects.

Promote and facilitate access to best practices in AI

The ia.anap.fr website references projects by category:

• Organization,
• Data management,
• Predictive medicine,
• Diagnostics,
• Patient follow-up,
• Decision making,
• Task automation.

It will be regularly updated and professionals will be able to easily view the history and maturity level of a project, the number of users who benefit from it and the key success factors identified.

Any professional can apply to Anap to submit a project and be referenced. After a selection committee has studied the project, the agency will be able to publish the solution in question on its platform.

During the press release announcing the launch of this platform, Anap published the following focus on four solutions referenced on its platform:

SurgAR, Pr. Nicolas Bourdel, Clermont-Ferrand University Hospital

The objective of the SurgAR project is to display in augmented reality the internal structure of organs during a surgical procedure performed by a minimally invasive approach (using a camera and small incisions). The organs thus become semi-transparent, the surgeon is truly guided in real time. This solution combining computer vision and artificial intelligence is intended to be deployed in operating rooms, regardless of the laparoscopic surgery equipment they have.

An online version is also available so that surgeons can train to use the tool. The entire project team is currently working on obtaining the CE mark.

Professor Nicolas Bourdel explains:

“Beyond the development of initiatives combining basic and clinical research that can help healthcare professionals, it is interesting to see to what extent these Artificial Intelligence projects should really also be exported as fully-fledged entities (startup / spinoff) outside of the hospital and university environment in which they were born. There is also an economic
There is therefore also an economic challenge, and more generally, a challenge in terms of how AI can create collaboration between hospitals, public research laboratories and private structures such as startups.

SUOG: Dr Ferdinand Dhombres, AP-HP

This project was born from the observation of a strong need for assistance in pregnancy ultrasound, linked in particular to the difficulties of access to experts and the complexity of possible diagnoses.

The solution proposed by SUOG is based on mixed artificial intelligence, a combination of machine learning (for ultrasound image recognition) and symbolic reasoning algorithms based on ontologies. The source data comes from 10 expert centers in Europe for the most exhaustive coverage of developmental anomalies listed by Orphanet, the portal for rare diseases.

The prospects of the project are numerous today. Large-scale clinical trials are planned from 2022 onwards, and the project should be commercialized by the end of 2023.

Dr. Ferdinand Dhombres states:

“With 130,000 cases of congenital anomalies and approximately 50,000 cases of ectopic pregnancy per year in Europe, the need for effective assistance during ultrasound screening has become critical. The SUOG assistant addresses the problem of access to expert sonographers and enables improved ultrasound screening for better organization of perinatal care.”

Know Your Patient, GHICL

The project is hosted by the Groupement des Hôpitaux de l’Institut Catholique de Lille (GHICL) and responds to the need to facilitate patient admission and improve patient care within the institution.

The solution is based on the recovery of algorithm models specific to the banking sector that automate the processing of identity documents, while providing a portal for admission agents. The project’s own prospects lie in the ongoing development of the third-party pathway (admission for a relative), the automatic recognition of the mutual card and the development of the solution’s integration with other hospital information system editors.

TransCUPtomics, Institut Curie

Faced with the difficulties of identifying the tissue origin of certain multi-metastatic cancers and while knowledge of the origin of a cancer is generally used as the basis for setting up an appropriate treatment, the Institut Curie has developed an artificial intelligence solution. Launched in 2019, this solution, which is now available to pathologists and oncologists, can identify the tissue origin of cancers of unknown primitive origin so that targeted treatment can be offered to the patient. During the evaluation and validation phases, TransCUPtomics will have identified 80% of cancerous tumors.

Discover all the projects listed on ia.anap.fr

Translated from L’Anap annonce le lancement d’une plateforme pour partager les solutions d’IA déployées dans les maisons de santé