Assistance Publique – Hôpitaux de Paris (AP-HP) announces the signature of a partnership with Gleamer. Gleamer is renowned for having designed a suite of AI software for radiology diagnosis. The firm will use its skills to co-build with AP-HP an AI solution to detect the main thoracic pathologies in standard radiography. This collaboration has one objective: to reduce the risk of diagnostic errors for patients received in emergency.

Succeeding in offering a quality preliminary diagnosis for patients received in emergency

Co-founded in 2017 in Paris by Christian Allouche, Alexis Ducarouge and Dr. Nor-eddine Regnard, Gleamer specializes in the development of AI software for radiology. Its suite of AI solutions is able to provide reliable preliminary diagnosis based on medical images and in particular on conventional radiography. Chest X-rays are often taken in an emergency situation and are sometimes subject to preliminary reading by non-radiologists.

AP-HP wishes to ensure that these images can be analysed by a high-performance tool that could accurately diagnose the main thoracic pathologies. This is why it has signed this partnership with Gleamer to design a solution that will ensure effective detection of pathologies. Marie Pierre Revel, Head of Radiology Department A at Cochin Hospital AP-HP, speaks about the importance of a quality solution:

“The possibility of having an algorithm that systematically reads, without bias, and in a reproducible manner regardless of the time and workload, all chest X-rays, is particularly interesting, to avoid detection errors, which are a source of delays in care: for example, a lung nodule that is not seen may reveal itself years later as an aggressive cancer.”

A partnership so that each party can contribute its expertise in designing a quality solution

Christian Allouche, CEO and co-founder of Gleamer commented on the signing of the partnership:

“We are delighted to partner with AP-HP which is recognized worldwide for its excellence in healthcare, research and teaching and which also has one of the most impressive pools of high-quality medical imaging data in the world! At GLEAMER, we already have artificial intelligence technologies deployed in many AP-HP centers, especially in osteoarticular.”

Concretely, AP-HP will allow the annotation of 100,000 thoracic radiographs illustrating pathologies of interest (pneumothorax, pleural effusion, alveolar syndrome, nodules, mediastinal masses, etc…) which will contribute to enrich and optimize the artificial intelligence algorithm developed by GLEAMER. Among the pathologies that the tool will be able to diagnose, we find lesions relevant to the emergency field such as pneumothorax, pleural effusion, infectious focus, pulmonary nodule or mediastinal mass.

Christian Allouche talks about the ambitions of his company:

“We are continuing our developments to improve and enrich our range of artificial intelligence software and have been working for several months with our thoracic imaging department which mobilizes dedicated artificial intelligence engineers, specialized radiologists, data scientists and researchers. Our ambition is to bring a simple and powerful solution, adapted to the needs of French healthcare institutions, to address priority public health issues.”

With this tool, Gleamer and AP-HP want to take pathology diagnosis a step further with AI.

Translated from AP-HP s’associe avec Gleamer pour innover dans la détection des anomalies radiologiques thoraciques