Artificial Intelligence, Atrial Fibrillation, Stroke
Conditions
Keywords
stroke, atrial fibrillation, artificial intelligence, risk prediction, electrocardiogram, non-vitamin K antagonist oral anticoagulants
Brief summary
We hypothesize that an AI-guided AF risk stratification approach, particularly when combined with intensified rhythm monitoring using wearable devices and extended ECG patches, will significantly increase AF detection rates compared with standard care. By enabling earlier identification of patients who may benefit from anticoagulation therapy, this strategy has the potential to improve clinical outcomes while minimizing unnecessary exposure to anticoagulant-related bleeding risks. Ultimately, this trial seeks to provide robust clinical evidence supporting the integration of AI-assisted ECG analysis into routine post-stroke care, advancing precision medicine and optimizing resource allocation for patients with ischemic stroke.
Detailed description
This study is designed as a multicenter, prospective, randomized controlled trial enrolling adult patients hospitalized for acute ischemic stroke who demonstrate sinus rhythm on admission 12-lead ECG. Participants will be randomized to either an AI-assisted care group or a usual-care control group, and further allocated to active or standard rhythm monitoring strategies. The primary endpoint is the incidence of newly diagnosed AF or atrial flutter within six months, defined by electrocardiographic documentation. Secondary and exploratory endpoints include the rate of oral anticoagulant initiation, AF burden metrics, recurrent embolic events, bleeding complications, major adverse cardiovascular events, cognitive outcomes, and all-cause mortality.
Interventions
DEVICEActive Follow-up Group
High-frequency electrocardiogram (ECG) monitoring was conducted, using a smartwatch with continuous photoplethysmography (PPG) monitoring, supplemented by an external single-lead ECG patch for 14 days. Participants recorded ECGs for 14 days immediately after discharge. If the smartwatch detected atrial fibrillation, the patch would be automatically applied for 14 days of recording. After recording, the patch was mailed or delivered in person to the research team for analysis.
OTHERStandard Follow-up Group
followed a standardized follow-up procedure by a neurologist. Each follow-up visit included a medical history and physical examination, without further cardiac-related monitoring. However, if the patient presented with symptoms of arrhythmia, the physician could consult a cardiologist or perform a preliminary 12-lead ECG.
Sponsors
National Taiwan University Hospital
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
DIAGNOSTIC
Masking
NONE
Intervention model description
A prospective randomized trial will enroll acute ischemic stroke patients with sinus rhythm at admission. Patients will be randomly assigned to either an intensive monitoring group, guided by the AI model, or a standard care group. High-risk patients identified by the AI model will undergo high density, 14-day external single-lead ECG monitoring, analyzed for AF episodes. Both groups will receive standard neurological follow-up. Primary outcomes include new-onset AF detection and recurrent stroke incidence, while secondary outcomes include AF burden, mortality, cognitive function, and healthcare resource utilization.
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
1. 18 years or older who are acutely hospitalized due to ischemic stroke; 2. A 12-lead electrocardiogram upon admission showing sinus rhythm;
Exclusion criteria
1. Previous diagnosis of atrial fibrillation or flutter; 2. Diagnosis of atrial fibrillation during hospitalization; 3. Pre-existing cardiac implantable electronic device (IVD); 4. Expected inability to attend regular outpatient follow-ups after discharge, or to undergo multiple 14-day ECG recordings; 5. Requires long-term anticoagulant use for other reasons, including but not limited to chronic pulmonary embolism; 6. Known contraindications to non-vitamin K antagonist oral anticoagulants due to comorbidities, including but not limited to end-stage renal disease, severe mitral stenosis due to rheumatic heart disease, or metallic heart valves; 7. Unwilling to sign a participant consent form.
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| The difference betwee Active Group & Standard Group | 400DAYS | The difference between the probability of a new diagnosis of atrial fibrillation or atrial flutter within six months (defined as an atrial fibrillation waveform longer than 30 seconds on a single-lead ECG or a twelve-lead ECG) in subjects identified as high-risk in the AI-assisted group and those who were actively followed up (Group A) and those who were followed up according to standard care (Group D). |
Countries
Taiwan
Contacts
CONTACTChih-Chieh Yu, MD.PhD
CONTACTHSIAO-HAN HUANG
Outcome results
None listed