At Centre Hospitalier de Condom, AZmed’s AI supports the first read of X-rays in the emergency department

PARIS, Oct. 12, 2026 -- AZmed, one of the leading AI companies in medical imaging, today announced that Centre Hospitalier de Condom has been using the full Rayvolve AI suite since 2022, primarily in its emergency department, where AI now supports the first read of trauma X-rays outside radiologist hours.

French emergency departments recorded 20.8 million visits in 2023, according to DREES, while the country has around 9,000 radiologists, or 14 per 100,000 inhabitants, with no demographic improvement expected for about ten years. In district hospitals, the emergency physician is often the first and only reader of those images, from the time the radiologist leaves in the evening until the following morning.

Rayvolve is CE-marked to help radiologists and emergency physicians detect and localize abnormalities on standard X-rays. Results are delivered with the images, with no separate login and no extra step in the workflow.

The AZtrauma module analyzes trauma X-rays for fractures, dislocations and joint effusions, as well as tumoral, degenerative and osteoarthritic lesions. A worldwide evaluation covered the full AZmed suite, 258,373 images from 100 centers across 26 countries; on the 195,706 musculoskeletal X-rays analyzed by AZtrauma, the module reached an area under the curve (AUC) of 0.983, with 97.4% sensitivity and 96.4% specificity. The AZchest, AZmeasure and AZboneage modules, also used at Condom and evaluated in the same study, cover thoracic findings, orthopedic measurements and skeletal maturity respectively.

“Outside regular hours, we do not often have a radiologist on site,” said Jérôme Bichon, Medical Imaging Manager at Centre Hospitalier de Condom. “The AI gives a second look at the first read done by our emergency physicians, even at two in the morning, and helps spot fractures.”

The department has formalized a protocol for uncertain cases: further imaging, including CT where indicated. The AI fits into that protocol without changing its indications.

“Vigilance drops on night shifts or during busy periods,” Bichon added. “The AI watches continuously and picks up the small fracture a tired physician might miss. After some initial reluctance, the emergency team adopted it widely. Physicians who have moved to other hospitals have asked their new employer for the same support.”

“The hardest test for our AI is the night shift, in a hospital where on-site radiology presence thins out. We are proud of the trust Centre Hospitalier de Condom has placed in us since 2022,” said Julien VIDAL, CEO of AZmed.

“At CH de Condom, it is the emergency physicians who move Rayvolve forward. Their QR code reports come to us straight from the field, and our teams use them to guide the next optimizations,” added Samuel Rizoulières, Account Manager France at AZmed.

Where overnight radiology cover no longer exists, AI is becoming part of how the first read is organized in emergency departments, in university hospitals and district hospitals alike.