A Dual-Branch Deep Learning Network Integrating Structured Data and Large Language Model-Based Medical Text Features for Atrial Fibrillation Recurrence Prediction: A Retrospective, Multicenter Study
Sponsor study identifier
observational
2,508
1 country
5
Study Start
April 15, 2025
Actual (reported)Primary Completion
June 30, 2025
Actual (reported)Study Completion
August 31, 2025
Actual (reported)We aimed to develop an dual-branch deep learning network that integrates feature representations extracted from medical texts using large language models, with structured data features derived from perioperative records of AF ablation patients. This integrated approach provides novel clinical insights into risk prognosis and enhances strategies for post-procedural management in AF ablation.
Registry-reported information. Elapsed time and enrollment totals alone do not measure study progress or recruitment pace.
August 31, 2025
Actual (reported)Click on a node to explore related trials.
Dates and qualifiers reported by the registry. Estimated dates do not confirm that a milestone has occurred.
April 15, 2025
Actual (reported)May 11, 2025
Actual (reported)May 18, 2025
Actual (reported)June 30, 2025
Actual (reported)August 31, 2025
Actual (reported)Late Atrial Fibrillation Recurrence
Recurrence occurring after 3 months are considered late recurrences.
Early Atrial Fibrillation Recurrence
Recurrences occurring within 3 months postoperatively are classified as early recurrences.
According to the latest ACC/AHA guidelines, AF recurrence is defined as the occurrence of atrial arrhythmias (atrial tachycardia, atrial flutter, or atrial fibrillation) lasting longer than 30 s following the procedure. Recurrences occurring within 3 months postoperatively are classified as early recurrences, while those occurring after 3 months are considered late recurrences.
We plan to extract data from patients diagnosed with atrial fibrillation who underwent their first atrial fibrillation ablation at the Fourth Affiliated Hospital of Zhejiang University School of Medicine, Taizhou Hospital of Zhejiang Province, the Affiliated Hospital of Yunnan University, Jinhua People's Hospital, and Beilun District Hospital. The collected data will include the patient's general information, laboratory tests, preoperative dynamic electrocardiogram text modality, echocardiogram text modality, and surgical procedure. The primary observation endpoint will be atrial fibrillation recurrence (defined as the detection of a persistent atrial arrhythmia lasting more than 30 seconds on ECG or dynamic ECG after ablation), with a follow-up period extending until March 2025.
The Affiliated Hospital of Yunnan University
Kunming, Yunnan, 650000, China
Jinhua People's Hospital
Jinhua, Zhejiang, 321000, China
Beilun District People's Hospital
Ningbo, Zhejiang, 315800, China
Taizhou Hospital of Zhejiang Province affiliated to Wenzhou Medical University
Taizhou, Zhejiang, 317000, China
The Fourth Affiliated Hospital of Zhejiang University School of Medicine
Yiwu, Zhejiang, 322000, China
2,508 participants (actual enrollment reported by the registry).
First Submitted
May 11, 2025
First Posted
May 18, 2025
Study Start
April 15, 2025
Actual (reported)Contact the study team to confirm eligibility.
Primary Completion
June 30, 2025
Actual (reported)Study Completion
August 31, 2025
Actual (reported)Last Updated
March 12, 2026
Record last verified: 2026-03
Download the complete ClinicalTrials.gov record imported by Syfrah, including fields not shown in this page. The registry may contain a newer revision.
Source: ClinicalTrials.gov.
JSON downloads include source attribution and retrieval details alongside the unchanged original record or the separately normalized values.
Source retrieved: