🔵 RESEARCH
AI Helps Marine Corpsmen Detect Lung Sliding on POCUS
A multi-reader study evaluated AI-assisted lung sliding detection in point-of-care ultrasound performed by Marine Corps corpsmen.
Demonstrates that AI decision support can compensate for limited ultrasound training in austere and prehospital settings, expanding who can reliably perform diagnostic lung ultrasound beyond credentialed sonographers.
Journal of special operations medicine : a peer reviewed journal for SOF medical professionals · 2026-03-16
Read more →
🔵 RESEARCH
Deep Learning Generates Synthetic Contrast-Enhanced Mammograms Without Iodine
Researchers used deep learning to synthesize iodine-enhanced mammograms from low-energy contrast-enhanced mammography images alone.
If validated, this approach could eliminate contrast media injection in CEM, removing allergic and renal risks while cutting procedure time and cost — a meaningful shift for breast imaging workflows.
European radiology experimental · 2026-03-16
Read more →
🔵 RESEARCH
DETECT-AS Trial Tests AI Biomarker for Aortic Stenosis Monitoring
A multi-site prognostic study will validate a deep learning biomarker score for risk stratification in older adults with aortic sclerosis or mild stenosis.
Mild aortic stenosis is common but inconsistently monitored. An AI-derived digital biomarker could personalize surveillance intervals and flag progression earlier, potentially changing echo follow-up protocols.
🔵 RESEARCH
Real-Time 3D Bone Reconstruction Achieved at Very-Low-Dose Levels
Researchers developed real-time 3D bone model reconstruction using very-low-dose imaging protocols.
Significant radiation reduction in skeletal imaging matters most for pediatric and repeat-scan orthopedic populations. Real-time capability could also enable intraoperative 3D guidance at dose levels approaching plain radiography.
🔵 RESEARCH
Multimodal LLMs Learn to Interpret 12-Lead ECG Images
Researchers trained multimodal large language models to comprehend and interpret 12-lead electrocardiographic images directly.
Enabling LLMs to read ECG images — not just structured data — opens a path to unified AI interpretation across imaging and diagnostic modalities, with implications for integrated cardiac reporting workflows.
🟢 FUNDING
Sectra Acquires Oxipit as RadNet Tops $300M in 2026 Deals
Sectra is acquiring AI firm Oxipit, RadNet has completed $300M in 2026 acquisitions, and Hologic's $18B PE sale clears a regulatory hurdle.
Three major consolidation moves in a single cycle. Sectra adds AI chest X-ray capability to its PACS platform, RadNet scales outpatient imaging assets, and Hologic's PE deal signals private capital's appetite for imaging at scale.
Google News: medical imaging AI acquisition · Mon, 16 Ma
Read more →
🩺 EDITOR’S TAKE — FOR CLINICIANS
AI decision support is pushing into austere and non-specialist settings — from military lung ultrasound to synthetic contrast mammography. Watch for these tools to reshape who can perform diagnostic imaging and where.
📊 EDITOR’S TAKE — FOR INVESTORS
Sectra's Oxipit acquisition and RadNet's $300M deal spree confirm that imaging AI consolidation is accelerating. Standalone AI companies face growing pressure to partner or sell as platform buyers expand capabilities.
🏭 EDITOR’S TAKE — FOR INDUSTRY
The convergence of ultra-low-dose reconstruction, synthetic contrast generation, and AI-assisted interpretation signals a workflow shift: less radiation, fewer consumables, and broader operator access. OEMs must build or buy accordingly.