Raidium, a Paris and Silicon Valley radiology startup, has introduced its AI-driven imaging platform, Raidium Read, at the Moffitt Cancer Center in the US. The platform replaces legacy tools and is currently used for clinical trials and research, with FDA clearance anticipated in 2026.
- Automates tumor measurements with AI foundation model
- Cuts variability in tumor tracking by three times
- Deployed at Moffitt Cancer Center, FDA approval expected 2026
What happened
Raidium, a startup operating out of Paris and Silicon Valley, launched its AI-native radiology platform named Raidium Read at the Moffitt Cancer Center, a leading oncology research institution in the US. This platform replaced Moffitt’s older radiomics applications and is now available for use in clinical trials and research projects. Raidium expects to secure FDA 510(k) clearance for the system before the end of 2026.
The system is built around Curia, Raidium’s proprietary AI foundation model that was trained on over 200 million CT and MRI slices from 150,000 exams. Rather than integrating AI as an add-on to existing medical imaging viewers, Raidium developed an entirely new viewer embedded with the model, designed to automate and improve tumor tracking workflows.
Why it matters
Oncology radiologists traditionally perform manual tracking of lesions across multiple imaging scans, a process that is time-consuming, labor-intensive, and subject to significant variability between readers. Raidium Read changes this by automatically detecting, segmenting, and measuring lesions across anatomical regions and mapping historical data against new scans, removing much of the manual burden.
This automation cuts inter-reader variability by a factor of three, improving consistency and reliability in monitoring tumor response to treatment. The innovative approach of building the viewer around the AI model rather than bolting AI tools on existing viewers promises faster deployment without backend integration, potentially accelerating the adoption of AI in clinical radiology workflows.
What to watch next
Further developments to monitor include how the platform integrates with existing hospital IT systems, its impact on radiologist productivity, and additional clinical studies demonstrating improved patient outcomes through AI-assisted radiology. Raidium’s success could serve as a blueprint for embedding AI models directly into medical imaging viewers across oncology and other specialties.