Foundation Model for Post-MI Cardiovascular Event Alert Using Huawei Watch

Tongji Hospital

Status: Not yet enrolling

Conditions: Coronary Artery Disease; Myocardial Infarction

Treatments: Device: Conventional care; Device: Smartwatch + AI assisted; Device: Smartwatch only

Study type: Interventional

Funder types: Other

Identifiers

NCT07395648

2025S128

Details and patient eligibility

About

The goal of this randomized controlled trial is to evaluate whether a HUAWEI smartwatch integrated with a foundation model can enhance daily monitoring and management of AMI patients during outpatient cardiac rehabilitation after PCI.

Can a HUAWEI smartwatch with a foundation model improve daily monitoring/management of post-PCI AMI patients during 3-month outpatient CR, vs routine management or smartwatch without AI? Does the AI-enabled smartwatch (Intervention I) reduce 3-month cardiovascular-related hospital/emergency visits vs the other two groups? Do smartwatches (with/without AI) enhance quality of life and activity function post 3-month CR? Does the AI-enabled smartwatch lower 1-year major adverse cardiovascular events (MACEs) and related hospital/emergency visits vs other groups? How accurate is the AI-enabled smartwatch's alert system (vs hospital diagnostic results) for cardiovascular abnormalities? What is the usability of the AI-enabled smartwatch for patients and doctors?

Participants will:

Sign informed consent; complete baseline exams (MET, SF-36, 6-minute walk, exercise tolerance test) and questionnaires (demographics, medical history, medications).

Receive 3-month outpatient CR. Attend follow-ups; report all adverse events promptly.

Intervention I (AI-Enabled Smartwatch) group Wear the smartwatch continuously for 3 months (sleep wear recommended); actively collect ≥30s ECG via the watch's crown.

Pause activity if heart rate exceeds anaerobic threshold; adjust exercise intensity if METs mismatch prescription.

Contact doctors immediately upon receiving abnormality alerts (e.g., heart failure); follow expert recommendations for further care; complete usability questionnaires.

Intervention II (Non-AI Smartwatch) group Wear the smartwatch continuously for 3 months (sleep wear recommended) to track activity/sleep, monitor real-time heart rate/MET.

Pause activity if heart rate exceeds anaerobic threshold; adjust exercise intensity if METs mismatch prescription.

Control Group Complete routine outpatient CR and follow-ups (no smartwatch use).

Enrollment: 1,050 estimated patients

Sex: All

Ages: 18 to 69 years old

Volunteers: No Healthy Volunteers

Inclusion criteria

1. Aged more than equal to 18 years, less than 69 years.
2. Discharged within 6 months after PCI for acute myocardial infarction, and haven't received outpatient cardiac rehabilitation;
3. 40%≤EF≤50% and NT-pro BNP ≤500 pg/mL;
4. With the history of hypertension;
5. Voluntarily joined and signed the informed consent.

Exclusion criteria

1. With contraindication for rehabilitation activities include: unstable angina, uncontrolled hypertension (systolic blood pressure>180 mmHg and/or diastolic blood pressure>110 mmHg in a quiet state), severe symptomatic heart valve stenosis, uncontrolled heart rate increase (>130 beats/minute), uncontrolled heart failure, third degree atrioventricular block (without pacemaker implantation), active pericarditis or myocarditis, presence of newly formed embolism (pulmonary or systemic circulation), aortic dissection, acute thrombophlebitis, acute systemic disease or fever, other metabolic abnormalities (such as acute thyroiditis, electrolyte imbalance, uncorrected blood glucose abnormalities or insufficient blood volume), and severe psychological disorders;
2. Patients who have a pre-planned staged revascularization procedure (e.g., planned second coronary intervention) following their initial PCI for AMI.
3. Tattoos or other substances that affect optical signals on the wrist.
4. Coronary heart disease patients at high risk.
5. Pregnant women or women planning to become pregnant within the next year.
6. Severe arrhythmia, including third degree atrioventricular block, ventricular escape rhythm, severe sinus bradycardia, sick sinus syndrome, supraventricular tachycardia, ventricular tachycardia, atrial fibrillation, atrial flutter, ventricular fibrillation, ventricular flutter.
7. Physical disability, blindness, and deafness.

Trial design

Primary purpose: Health Services Research

Allocation: Randomized

Interventional model: Parallel Assignment

Masking: Single Blind

---

1,050 participants in 3 patient groups

Intervention I group (the smartwatch with AI assisted group)

Experimental group

Description:

Subjects in the intervention group will be required to wear a smartwatch continuously for 3 months. For patients in Intervention group, the smartwatch will record the daily activities and sleep status. Furthermore, for patients in Intervention I group, the pre-trained foundation model will generate preliminary diagnostic suggestion based on the data collected by the smartwatch to remind the appearance of heart failure, myocardial ischemia, arrhythmia, abnormal blood pressure, and abnormal cerebral blood supply.

Treatment:

Device: Smartwatch + AI assisted

Intervention II group (smartwatch assisted group)

Experimental group

Description:

Subjects in the intervention group will be required to wear a smartwatch continuously for 3 months. For patients in Intervention group, the smartwatch will record the daily activities and sleep status.

Treatment:

Device: Smartwatch only

Control group ( conventional care)

Active Comparator group

Description:

Participants will receive routine outpatient CR and follow-up.

Treatment:

Device: Conventional care