Weekly AI digest
Radiology & medical imaging AI · week of 5 to 11 May 2026
6 peer-reviewed papers · 4 industry and regulatory items · conference and KOL highlights. Adapted from my weekly intelligence report. Full report (PDF, FR).
Three things that mattered
Opportunistic imaging scales to population cohorts
Two large studies on the NAKO cohort show the reach of automated analysis: AI body composition on 66,608 whole-body MRI scans (Radiology) and deep-learning detection of sacroiliac bone marrow edema in 11,163 participants (The Lancet Rheumatology), where AI-detected prevalence reached 30.8%, well above current clinical estimates.AI governance becomes standardized
On 5 May, the ACR and SIIM jointly published a practice parameter setting minimum requirements for responsible clinical deployment of imaging AI, covering pre-deployment validation, real-world monitoring, and bias management. The ACR also launched Assess-AI, a national post-market performance registry for radiology algorithms.Commercial access models raise equity questions
SimonMed Imaging introduced paid AI analysis modules offered as a direct patient option, including AI mammography review and coronary calcium detection. The move opened a debate on equitable access to diagnostic AI tools and on the medico-legal responsibilities involved.
Peer-reviewed
AI for automated body composition analysis on whole-body MRI in a large population cohort (n = 66,608)
Radiology (RSNA) · University Medical Center Freiburg, May 2026Deep-learning detection of sacroiliac bone marrow edema on MRI: prevalence and clinical correlates in 11,163 NAKO participants
The Lancet Rheumatology · Charite University Hospital Berlin, 6 May 2026ResNet automation of the field of view in prostate MRI: prospective clinical validation
Scientific Reports · University Medical Center Dusseldorf, 11 May 2026Deep-learning automated segmentation of multiple myeloma osteolytic lesions on whole-body CT
Computers in Biology and Medicine · Leiden University Medical Center, 6 May 2026Computer-aided diagnosis of age-related macular degeneration: above 97% accuracy with multimodal deep learning
Scientific Reports · Mansoura University, 9 May 2026PrecisionView: a deep-learning endomicroscope for real-time high-resolution tissue imaging during endoscopy
PNAS · University of California San Diego, 11 May 2026
Industry & regulation
RIVANNA Medical clears Accuro XV via FDA 510(k), an AI-guided ultrasound for regional anesthesia
7 May 2026. Portable ultrasound with automatic identification of nerve and bone structures to guide regional nerve blocks, reducing procedure time and operator dependence.Viz.ai earns ISO/IEC 42001 certification for AI management systems
5 May 2026. First medical AI company to reach this global AI governance standard (ISO/IEC 42001:2023), positioning ahead of EU AI Act requirements.Azra AI launches an agentic clinical research platform for oncology
6 May 2026. Uses oncology-tuned LLMs and autonomous agents to automate patient recruitment for clinical trials, designed to integrate with Epic and Oracle Health EHRs.SimonMed Imaging introduces paid AI analysis modules as a direct patient option
5 to 7 May 2026. Add-ons include AI mammography analysis and coronary calcification detection on chest CT, raising ethical and medico-legal questions on equitable access.
Conferences & voices
The week centered on governance. The joint ACR-SIIM practice parameter, published 5 May, was framed by working group chair Tessa Cook (University of Pennsylvania) as a milestone for responsible deployment, giving institutions a clear framework for pre-launch validation and post-launch monitoring. SIIM board chair Nabile Safdar (Emory University) positioned radiology as the field to lead responsible AI implementation in health care. On the new Assess-AI registry, ACR board member Christoph Wald (Lahey Hospital & Medical Center) noted that purchasing AI is now straightforward while knowing whether it works for a given practice and payer mix is the hard part.
Source articles are indexed in PubMed with verified DOIs. Manuscript-stage work is excluded. The full weekly report is produced in French.
Get the weekly digest
One email a week on what matters in radiology and imaging AI. Researched by a practicing radiologist. Free, no vendor sponsorship.