The practical stake is a first-day clinical decision: which burn areas are likely to heal on their own, and which may need significant intervention. Spectral AI says its DeepView System has received FDA De Novo Classification, giving the company authorization to begin US commercial distribution for the burn indication.

According to the company’s release, DeepView is intended for burn care settings including burn centers, trauma centers, and emergency departments. That matters because burn triage is time-sensitive: early judgment shapes transfer decisions, surgical planning, length of stay, and cost.

The system is non-invasive and combines multispectral imaging with a proprietary AI algorithm. Spectral AI says image acquisition takes 0.2 seconds, while image processing and AI model classification take about 20 to 25 seconds. The clinical output is designed to indicate whether burn-wound areas are unlikely to heal within 21 days and may require more intensive medical intervention.

The training base is also notable. The company says DeepView was trained and tested against a proprietary, clinically validated database of more than 340 billion pixels of burn-wound image data. For healthcare AI, that kind of claim only becomes commercially meaningful when it survives regulatory review and fits into real clinical workflow. De Novo clearance is therefore more than a marketing milestone: it creates a US pathway for a device category without relying on a direct predicate device.

The next question is adoption. Hospitals will still need to evaluate procurement, workflow integration, clinician trust, and reimbursement. But the clearance pushes AI-assisted wound assessment from validation story to commercial test. If DeepView performs consistently outside trial conditions, burn care could become one of the clearer examples of AI improving a narrow, high-stakes medical decision rather than trying to replace a broad clinical role.

#deepview

DeepView clears FDA De Novo as burn-care AI heads for US rollout