Back to Tracker

Beijing Tiantan Hospital

China Healthcare

AI-generated radiology report draft with doctor review

Automation Workflow Redesign

Measured outcomes Evidence code E4Scaled use

What changed

What changed

AI generates an initial head-CT report draft that is then reviewed and finalized by a doctor.

Key results

Key results

AI report-generation time — Report generated in less than one minute — After 1 minutes

2026

Reported diagnostic accuracy — Approximately 80% diagnostic accuracy — After 80 percent

2026

How the organization responded

How the organization responded

Organizational response: Not yet evidenced

Why it matters

AIGW analysis

Radiology documentation is partially automated while final review remains human. Human verification is an explicit workflow control, but the source does not establish a changed capability requirement; Judgment & Verification is therefore an implication, not an observed capability shift.

Capability implications: Judgment & Verification

Human role / decision rights

Human role and decision rights

AI writes the first draft; a doctor reviews and finalizes it.

Source

Source & evidence

How we verify evidence

Digital China

AI-generated head-CT report drafts at Beijing Tiantan Hospital

Government or public institution

Government-hosted article, attributed to People's Daily Overseas Edition, reports AI-generated head-CT report drafts, doctor review/finalization, sub-one-minute generation and approximately 80% diagnostic accuracy.