Artificial Intelligence-assisted Diagnosis

Artificial Intelligence-assisted Diagnosis of Congenital Heart Diseases on Chest X-ray

This study aims to investigate whether an AI-assisted decision support can improve the diagnosis of congenital heart diseases on Chest X-ray.

Congenital heart disease (CHD) is a structural abnormality of the heart and/or great vessels and the frequency of CHD is nearly 1% in general population. In clinical practice, Chest X-ray is an important routine method for detection of CHD. However, its learning curve is high. Based on Chest X-ray, Artificial intelligence (AI) has the potential to improve the efficiency and accuracy of CHD. This study aims to investigate whether an AI-assisted decision support can improve the diagnosis of congenital heart diseases on Chest X-ray.

This study will enroll subjects with CHD, and the enrollment will be divided into two phases.

Phase 1: Previous data of chest X-ray in patients with CHD. A minimum of 5000 patients is reached.

Phase 2: Prospective inclusion of general patients with chest X-ray. A minimum of 50000 patients is reached.

Inclusion Criteria:

1. Chest X-ray of patients with ASD;
2. Chest X-ray of patients with VSD;
3. Chest X-ray of patients with PDA;
4. Chest X-ray of patients with PFO;
5. Chest X-ray of patients with PS;
6. Chest X-ray of patients with TOF/PAA;
7. Chest X-ray of patients with MS;
8. Chest X-ray of patients with other heart diseases;

Exclusion Criteria:

1. CHD Patients without surgical, interventional validation;
2. Serious Scoliosis;
3. Pulmonary inflammation or consolidation