Skip to content

Construction of Atrial Fibrillation Prediction Model Based on Sinus Rhythm Electrocardiograms Using Deep Learning

Construction of Atrial Fibrillation Prediction Model Based on Sinus Rhythm Electrocardiograms Using Deep Learning

Status
Active, not recruiting
Phases
Unknown
Study type
Observational
Source
ChiCTR
Registry ID
ChiCTR2500110667
Enrollment
Unknown
Registered
2025-10-17
Start date
2025-10-20
Completion date
Unknown
Last updated
2025-10-20

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

atrial fibrillation ,AF

Interventions

AF low risk group:NA
AF high risk group:NA
Retrospective study observation group:NA

Sponsors

Guangdong Second Provincial General Hospital (Guangdong Provincial Emergency Hospital)
Lead Sponsor

Eligibility

Sex/Gender
All
Age
18 Years to No maximum

Inclusion criteria

Inclusion criteria: 1. Retrospective study: The patients who visited our hospital and underwent electrocardiogram (ECG) more than twice from January 2015 to May 2023 were selected as the research subjects. According to whether atrial fibrillation was detected in the second examination, they were classified into two categories: the normal group and the atrial fibrillation group. The ECGs of the patients' first examination, that is, the electrocardiograms of sinus rhythm, were collected to construct an ECG database. 2. Prospective study: The electrocardiograms collected in the Cardiac Function Room of our hospital from April 2024 to June 2024 were collected. Based on the improved AlexNet model, AI electrocardiograms were formed. According to the output values of the AI electrocardiograms, the subjects were divided into the low atrial fibrillation (AF) risk group and the high AF risk group, with 100 people included in each group.

Exclusion criteria

Exclusion criteria: 1. If the electrocardiogram examination shows atrial fibrillation or atrial flutter for the first time or if the quality of the graphic acquisition is not good, the subjects will not be included in the study. 2. Prospective study: Patients will not be included in the research subjects if they are diagnosed with atrial fibrillation/atrial flutter during the electrocardiogram (ECG) examination; if the quality of the graphic acquisition is poor; if they are unable to participate in the clinical study; or if they refuse to sign the informed consent form.

Design outcomes

Primary

MeasureTime frame
Accuracy ;Precision;F1-Score;Recall;Sensitivity;Specificity;

Countries

China

Contacts

Public ContactHuang Jinghe

Guangdong Second Provincial General Hospital (Guangdong Provincial Emergency Hospital)

953825060@qq.com+86 20 8916 8235

Outcome results

None listed

Source: ChiCTR (via WHO ICTRP) · Data processed: Feb 4, 2026