Trial of Artificial Intelligence for Chest Radiography (ACER)

Duke University

Status: Not yet enrolling

Conditions: Lung Cancer; Pneumonia

Treatments: Diagnostic Test: AI

Study type: Interventional

Funder types: Other

Identifiers

NCT06456203

2023/2280

Details and patient eligibility

About

Randomized Clinical Trial of the impact of Chest radiograph AI-assisted triage and report generation upon clinical outcomes and an economic analysis of impact of AI decision support on radiology service delivery.

Full description

Randomized, prospective selection of patients. Control group involves radiologists reporting chest radiographs as per reference standard clinical workflow Intervention group involves radiologists assisted with AI reporting an AI-triaged worklist of chest radiographs using an AI report generation tool Clinical outcomes on patients are studied at pre-determined study endpoints, including time to discharge from the hospital and re-admission rates.

Economic analysis on cost-avoidance from man-hours saved from report generation and triage.

Enrollment: 10,000 estimated patients

Sex: All

Ages: 14 to 130 years old

Volunteers: Accepts Healthy Volunteers

Inclusion criteria

• All patients attending radiography to have chest radiographs during the study period

Exclusion criteria

• age below 14
• deceased before discharge
• chest radiograph performed in non-standard projections

Trial design

Primary purpose: Diagnostic

Allocation: Randomized

Interventional model: Parallel Assignment

Masking: None (Open label)

---

10,000 participants in 2 patient groups

Control Arm

No Intervention group

Description:

Chest radiographs reported with AI assistance

AI assisted

Active Comparator group

Description:

AI assisted detection, triage and reporting of CXR

Treatment:

Diagnostic Test: AI