DigitHeart Echo Study (DigitHeart-SHD)

Cardiac Anatomical and Mechanical Properties Prediction From Electrocardiography (ECG) With Multi-Modal Representation Learning

Eligible patients will be interviewed by research staff to explain the trial design and rationale. Written informed consent will be obtained from patients who voluntarily agree to participate. Demographics and medical history will be obtained. A 12-lead ECG will be performed; it has not been performed within 6 months. A photo of the 12-lead ECG will be taken by research staff using the DigitHeart-2 smartphone application. Predictions from DigitHeart-2/MERL-ECHO and other machine learning models hosted by the investigator team for each of the target anatomical and mechanical properties will be recorded. The cardiologist in charge of the patient, who is blinded to the prediction results from DigitHeart-2/MERL-ECHO, will perform echocardiography according to American Society of Echocardiography guidelines, which will serve as the gold standard for accuracy evaluation.

Study Overview

• Status: Recruiting
• Conditions: Heart Disease; Echocardiography; Electrocardiograph
• Intervention / Treatment: Diagnostic test: echocardiography

• Study Type: Observational
• Enrollment (Estimated): 478

Contacts and Locations

• Study Contact: Name: Chun Ka Dr Wong, Clinical Assistant Professor; Phone Number: +852-22553111; Email: wongeck@hku.hk

Study Locations

• Hong Kong
  • Hong Kong, Hong Kong
    • Recruiting
    • Department of Medicine Queen Marry Hospital, Hong Kong
    • Contact: Chun Ka Dr Wong, Clinical Assistant Professor; Phone Number: +852-22553111; Email: wongeck@hku.hk
    • Principal Investigator: Chun Ka Dr Wong, Clinical Assistant Professor

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: Yes

• Sampling Method: Non-Probability Sample

Study Population

Adult patients aged ≥ 18 years old who fulfill recruitment criteria will be recruited from Queen Mary Hospital and Tung Wah Hospital, Hong Kong.

Description

Inclusion Criteria:

• aged ≥ 18 years old
• planned to have echocardiography performed
• voluntarily agree to participate in the trial.

Exclusion Criteria:

• had echocardiography performed within 1 month
• pacemaker rhythm on ECG
• dextrocardia
• complex adult congenital heart disease
• ventricular assist device implantation.

Study Plan

How is the study designed?

Number of groups / cohorts

1

Cohorts and Interventions

Group / Cohort

Intervention / Treatment

Smartphone-Captured ECG Cohort; 12-lead ECG will be performed if it has not been performed within 6 month. Photo of the 12-lead ECG will be taken by research staff using DigitHeart-2 smartphone application. Prediction from DigitHeart-2/MERL-ECHO and other machine learning model(s) hosted by the investigator team for each of the target anatomical and mechanical properties will be recorded. Cardiologist in-charge of the patient, who is blinded to the prediction results from DigitHeart-2/MERL-ECHO, will perform echocardiography according to American Society of Echocardiography guidelines, which serve as gold standard for accuracy evaluation.

Diagnostic test: echocardiography; Participants will undergo an echocardiography to determine their cardiac structure, heart function and valve function. The gold-standard measurement from the echocardiography text report will be the reference for validating the multi-modal prediction.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Model's Accuracy in predicting key echocardiographic parameters	The primary outcome is the accuracy of the integrated DigitHeart- 2/MERL-ECHO system in predicting key echocardiographic parameters from smartphone-captured ECG images, as measured by AUROC and F1 scores.	Day 1

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Model's Accuracy in predicting other key anatomical and mechanical properties	Secondary outcomes include accuracy of predicting other anatomical and mechanical properties including: Impaired left ventricular systolic function; Dilated right ventricle; Impaired right ventricular function; Moderate to severe aortic stenosis; Moderate to severe aortic regurgitation; Moderate to severe mitral stenosis; Moderate to severe mitral regurgitation; Moderate to severe tricuspid regurgitation	Day 1

Collaborators and Investigators

• Sponsor: The University of Hong Kong

Publications and helpful links

General Publications

• Wong CK, Lau YM, Lui HW, Chan WF, San WC, Zhou M, Cheng Y, Huang D, Lai WH, Lau YM, Siu CW. Automatic detection of cardiac conditions from photos of electrocardiogram captured by smartphones. Heart. 2024 Aug 14;110(17):1074-1082. doi: 10.1136/heartjnl-2023-323822.

Study record dates

Study Major Dates

• Study Start (Actual): February 14, 2025
• Primary Completion (Estimated): February 14, 2027
• Study Completion (Estimated): February 14, 2027

Study Registration Dates

• First Submitted: March 26, 2025
• First Submitted That Met QC Criteria: April 22, 2026
• First Posted (Actual): April 30, 2026

Study Record Updates

• Last Update Posted (Actual): April 30, 2026
• Last Update Submitted That Met QC Criteria: April 22, 2026
• Last Verified: April 1, 2026

More Information

Terms related to this study

• Additional Relevant MeSH Terms: Cardiovascular Diseases; Heart Diseases; Diagnostic Techniques and Procedures; Diagnosis; Diagnostic Imaging; Diagnostic Techniques, Cardiovascular; Heart Function Tests; Cardiac Imaging Techniques; Ultrasonography; Echocardiography
• Other Study ID Numbers: DigitHeart-SHD; HKWC/HKU IRB (Other Identifier: HKU/HA HKW Institutional Review Board)

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: NO

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No