AI Facial Analysis Algorithm to Screening Coronary Artery Disease in High-Risk Community Population

National Center for Cardiovascular Diseases

Status: Enrolling

Conditions: Screening; Artificial Intelligence Algorithms; Face; Coronary Artery Disease (CAD)

Treatments: Other: No Intervention

Study type: Observational

Funder types: Other

Identifiers

NCT07277686

2025-ZX165

Details and patient eligibility

About

This study aims to evaluate the effectiveness of this facial image-based AI algorithm for screening CAD in high-risk community populations (specifically individuals with diabetes, hypertension, or aged over 65). The main objectives are:

1. To verify if the AI algorithm can accurately distinguish between high-risk and low-risk groups by comparing the actual prevalence of CAD in these groups.
2. To compare the CAD detection rate using this AI screening strategy against the natural detection rate in a real-world cohort.

Enrollment: 1,392 estimated patients

Sex: All

Ages: 18+ years old

Volunteers: Accepts Healthy Volunteers

Inclusion criteria

• Age >= 18 years.

• Community high-risk population, defined as individuals meeting at least one of the following criteria:

  1. Diagnosed with Diabetes Mellitus;
  2. Diagnosed with Hypertension;
  3. Advanced age (> 65 years old).

Exclusion criteria

• Prior confirmed diagnosis of Coronary Heart Disease (CAD), including clinically diagnosed CAD, history of Coronary Artery Bypass Grafting (CABG), or history of Percutaneous Coronary Intervention (PCI).
• History of diagnosed Heart Failure.
• Significant facial alterations or conditions that may interfere with AI image analysis, such as plastic surgery, severe facial trauma, or heavy makeup.
• Refusal to participate in the study.

Trial design

1,392 participants in 2 patient groups

Diagnostic Test Cohort

Description:

Participants in this cohort undergo facial image-based AI screening. Subsequently, all participants in this group will undergo Coronary Computed Tomography Angiography (CCTA) as the gold standard reference to verify the diagnosis of CAD. All participants are followed up for 6 months.

Treatment:

Other: No Intervention

Real-world Observational Cohort

Description:

Participants in this cohort represent a real-world setting. They are followed up for 6 months to observe the natural detection rate of CAD and the occurrence of MACE outcomes.

Treatment:

Other: No Intervention