SimonMed highlights role in large global X-ray AI study

SimonMed has highlighted its role in a large peer-reviewed evaluation of AI for X-ray imaging published in Radiography. The study analyzed 258,373 X-rays from 100 medical centers across 26 countries and 5 continents, according to SimonMed.

Dr. Sean Raj, MD, chief medical officer and chief innovation officer at SimonMed, served as senior author. SimonMed said it contributed U.S. clinical data and played a role in study design, execution, and analysis.

Researchers evaluated the 4 components of AZmed’s Rayvolve AI Suite as a unified clinical platform under real-world conditions. SimonMed said the evaluation did not exclude exams based on image quality or acquisition protocol.

AZtrauma achieved an AUC of 98.3%, with sensitivity of 97.4% and specificity of 96.4%, across 195,706 musculoskeletal examinations. AZchest achieved an AUC of 97.8%, with sensitivity of 96.7% and specificity of 87.9%, across 61,418 chest radiographs covering 6 pathology categories.

The suite also included AZmeasure and AZboneage. SimonMed reported measurement precision within 1.83° for angles, 1.1 mm for lengths, and a bone age estimation error of about 6 months.

All 258,373 images were processed without a technical failure, according to SimonMed. The company said performance remained consistent across pathologies, anatomies, patient demographics, and clinical settings.

The collaboration builds on SimonMed’s 2023 selection of AZmed as its AI partner for X-ray diagnostics. SimonMed said the earlier deployment showed a 6x reduction in turnaround time for fracture cases and 98.5% sensitivity across its centers.

“This study represents a defining moment for AI in medical imaging,” Dr. Raj said. He said the work reinforces the importance of rigorous real-world clinical evidence.

AZmed’s scientific evidence page lists the study as “Performance of a complete AI radiographic suite across 258,373 X-rays from 26 countries: A worldwide evaluation.” The page describes it as a large-scale, international, multicenter retrospective evaluation of a complete radiographic AI suite across clinical settings.