This research project aims to develop and validate a tool that uses artificial intelligence (AI) to automatically detect and quantify aortic regurgitation (AR). The clinical efficacy of this tool will be established by comparing it to manual diagnostic methods in a multicenter randomized controlled trial. By leveraging deep learning (DL) techniques, the AI system will automate aortic regurgitation (AR) detection, measurement, and diagnosis, addressing challenges like variability in echocardiographic interpretations and the need for specialized expertise. It will integrate multiple echocardiographic parameters to provide accurate, standardized, and efficient AR diagnoses, reducing human error and improving consistency. This tool will enhance diagnostic precision and accessibility, improving clinical outcomes and extending advanced diagnostic capabilities to a broader range of healthcare environments, including resource-limited settings.
Study Type
INTERVENTIONAL
Allocation
RANDOMIZED
Purpose
DIAGNOSTIC
Masking
DOUBLE
Enrollment
540
Participants in this group will undergo aortic regurgitation assessment using an advanced artificial intelligence tool.
Participants in this group will receive a traditional diagnostic assessment for aortic regurgitation, performed by trained sonographers following standard protocols.
Division of Cardiology, Department of Medicine and Therapeutics Faculty of Medicine, The Chinese University of Hong Kong
Hong Kong, New Territories, Hong Kong
RECRUITINGStudy Outcomes
To compare the accuracy of the AI group and the manual group in distinguishing severe from non-severe AR, using expert cardiologists' (ASE level III or equivalent) assessments as the reference standard.
Time frame: This will be recorded from baseline to study completion (20 months)
Comparing Accuracy in Differentiating AR Severity Levels
To compare the accuracy of the AI group and the manual group in differentiating trace, mild, moderate, and severe aortic regurgitation, using cardiologists' assessments as the reference standard.
Time frame: This will be recorded from baseline to study completion (20 months)
Assessing deviations in Effective Regurgitant Orifice Area (EROA)
The Effective Regurgitant Orifice Area (EROA) assesses the size of a valve opening that leads to backward blood flow in the heart. It is an important metric for evaluating valvular regurgitation and will be measured during echocardiography.
Time frame: This will be recorded from baseline to study completion (20 months)
Assessing deviations in Vena Contracta (VC)
The Vena Contracta (VC) is an important measurement in echocardiography used to evaluate how severe mitral regurgitation is and will be measured during echocardiography.
Time frame: This will be recorded from baseline to study completion (20 months)
Assessing deviations in Proximal Isovelocity Surface Area (PISA)
Proximal Isovelocity Surface Area (PISA) is a method used in echocardiography to evaluate mitral regurgitation and will be measured during echocardiography.
Time frame: This will be recorded from baseline to study completion (20 months)
Assessing deviations in jet width
The jet width is a critical measurement used to assess the severity of aortic regurgitation and will be measured during echocardiography.
Time frame: This will be recorded from baseline to study completion (20 months)
Assessing deviations in Regurgitant Volume (RegVol)
Regurgitant Volume assesses how much blood leaks back into the left atrium during mitral regurgitation and will be measured using Doppler echocardiography.
Time frame: This will be recorded from baseline to study completion (20 months)
Comparing Assessment Completion Time
To compare the time taken by the AI group, the manual group, and the cardiologists to complete their assessments.
Time frame: The time taken for each method to reach a diagnosis will be recorded from baseline to study completion (20 months)
Tracking 1-Year Outcomes
To track 1-year all-cause mortality and heart failure hospitalizations (HFH), comparing outcomes for patients with severe aortic regurgitation identified by the AI and manual groups, separately.
Time frame: Participants will be followed up at 6 and 12 months to monitor outcomes, including 1-year all-cause mortality and HFH.
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.