AI CT reporting is not a detection overlay or an alert system. It is a complete reporting service: your DICOM series goes in, a ready-to-sign preliminary report comes out — produced by two AI layers and reviewed by our in-house radiologist
worklist prioritizers
No complete report is delivered
AI CT reporting
Complete ready-to-sign preliminary report delivered
Every xAID study follows this exact workflow — 24/7, including after-hours and weekends, at no surcharge
Your PACS pushes the CT series to xAID via standard DICOM C-STORE — the same protocol used to send studies to any teleradiology provider. No custom middleware, no manual uploads. The study arrives and queues for AI processing automatically
The first AI layer — a foundation model designed for volumetric CT analysis — analyzes the full DICOM series. It identifies 100+ potential findings across anatomical structures: lung nodules, coronary calcium, stroke signs, hemorrhage, organ findings, fractures, vascular measurements
A second AI layer cross-checks the foundation model's findings for consistency and completeness. This dual-verification step catches false positives and ensures no significant finding is missed. The output is a structured, finding-by-finding report draft
Our in-house radiologist reviews the complete AI-generated report against the DICOM series. They can modify, add, or remove findings. Once satisfied, they approve the report for delivery
The preliminary report is returned to your RIS via HL7 ORU message — or as PDF if preferred. It appears in your system exactly as a teleradiology report would. Turnaround 2–12 hours from study receipt, 24/7 at flat per-study rate
Volumetric foundation model designed for 3D CT interpretation — not a repurposed 2D medical imaging model. Analyzes the full CT series across 100+ finding categories with high accuracy verified by peer-reviewed studies
A secondary model cross-verifies the foundation model output for consistency and completeness. Catches potential missed findings and reduces false positives before the report reaches the radiologist
A radiologist reviews the complete AI-generated report against the original DICOM series. Any finding can be modified, added, or removed. The radiologist is the accountable physician for every report
Understanding the three categories helps you evaluate the right solution for your workflow
| Category | Output | Radiologist reports? | After-hours surcharge | Quality guarantee |
|---|---|---|---|---|
| AI Detection Tools (worklist prioritizers) | Alert or worklist flag | No — radiologist still reads from scratch | N/A — continuous monitoring | None |
| Traditional Teleradiology | Signed CT report | Yes — dictated manually | 30–100% surcharge | None |
| AI CT Reporting (xAID) | Ready-to-sign preliminary report | Yes — AI generates, radiologist reviews | Zero — flat rate 24/7 | 95% accuracy verified |
xAID achieves 95% accuracy verified by independent peer-reviewed studies — not internal benchmarks. Full methodology, study details, and results are available on the clinical evidence page
AI CT reporting is a service that delivers complete ready-to-sign preliminary reports using AI, with radiologist review on every study. "AI radiology" covers two very different products: (1) AI detection tools, which flag suspected high-acuity findings and reprioritize the worklist but do not produce a complete report; and (2) AI CT reporting services like xAID, which generate the complete structured preliminary report using AI — with a radiologist reviewing before delivery. The key question is: does the AI produce a complete preliminary report, or just an alert? xAID produces the complete preliminary report
xAID does not replace the radiologist — it changes their role. Instead of spending 15–20 minutes dictating a report from a blank page, the radiologist reviews a complete AI-generated structured report in 3–5 minutes and delivers it. The radiologist is still the accountable physician for every report. The AI does not act autonomously; it generates the draft that the radiologist reviews. This human-in-the-loop model is why xAID is HIPAA compliant and achieves 95% accuracy verified by peer-reviewed studies
xAID achieves 95% accuracy verified by independent peer-reviewed studies. These figures reflect AI-plus-radiologist performance, not autonomous AI, since every xAID report is reviewed by our in-house European radiologist before delivery. Full clinical evidence is available at xaid.ai/accuracy
xAID uses a volumetric foundation model designed specifically for 3D CT interpretation, not a repurposed 2D model. A secondary AI layer cross-checks the primary model's output for consistency before the radiologist review stage. This two-layer verification underpins the 95% accuracy verified by peer-reviewed studies
From DICOM receipt to report delivery: 2–12 hours. This covers AI processing, radiologist review queue, and report delivery. Configurable SLAs are available for urgent studies. The 2–12 hour range applies 24/7 — nights, weekends, and holidays at no additional cost. The ACR guideline for routine CT turnaround is 24 hours; xAID typically delivers well within that window
Free 5-study pilot. Send CT studies from your actual case mix — head, chest, or abdomen. Receive preliminary reports within 12 hours. No integration or commitment required to start