12/04/2025
Something big is happening at RSNA 2025 — and it’s reshaping the future of radiology.
For years, radiology AI has relied on specialized models that each solve one small problem. But this year, the momentum has shifted toward foundation models that can understand many types of medical images, reason across context, and support radiologists through full workflows.
We’re seeing:
• Foundation models outperforming radiology trainees
• Faster reporting across huge imaging datasets
• Radiologists packing classrooms to learn agentic AI
• Real evidence that broad, generalist models are more powerful and practical than the old “model per organ” approach
At Braid, we’ve been testing and adopting every major foundation model into our Ask AI tool — including the newest jump to Gemini 3.0 — and the difference is real.
We break down the shift happening at RSNA here:
👉 https://braid.health/www/page/rsna-2025-foundation-models-radiology-ai
If you care about radiology, AI, or the future of healthcare, this moment matters.