This single-center, prospective, observational cohort study aims to evaluate the clinical application value of multi-modal cardiovascular magnetic resonance (CMR) imaging in patients with cardiovascular diseases (CVD).
While traditional imaging methods have limitations in fully evaluating myocardial tissue characteristics, multi-modal CMR offers a comprehensive, non-invasive "one-stop" assessment. It can simultaneously evaluate heart structure, function, tissue features (such as fibrosis and edema), and hemodynamics.
The study plans to enroll patients with suspected or confirmed CVD. Participants will undergo a comprehensive multi-modal CMR scan (including Cine, T1/T2 mapping, Late Gadolinium Enhancement, and 4D flow sequences) as part of their evaluation. In addition to clinical evaluation, the study will explore sequence optimization (e.g., comparing pre-contrast vs. post-contrast Cine, and 3-slice vs. 9-slice T1 mapping) and evaluate deep learning-based virtual native enhancement (VNE) models trained under different sequence protocols.
By tracking clinical outcomes, the study seeks to establish a standardized imaging assessment system to improve the early detection, accurate diagnosis, risk stratification, and prognostic prediction for various types of cardiovascular diseases.
Aged 18 years and older, with no gender restrictions.
Absolute contraindications: Implantation of non-MRI compatible metallic foreign bodies (e.g., old pacemakers, implantable cardioverter-defibrillators [ICD], aneurysm clips, etc.).
The study population consists of consecutive adult patients (aged 18 years and older) who are clinically suspected or confirmed to have cardiovascular diseases (CVD), including but not limited to ischemic heart disease, non-ischemic cardiomyopathy, myocarditis, and valvular heart disease. These participants are enrolled from the Second Affiliated Hospital Zhejiang University School of Medicine. They are patients who are scheduled to undergo a multi-modal cardiac magnetic resonance (CMR) examination as part of their clinical evaluation to determine disease etiology or assess myocardial tissue characteristics.
Cohort
Prospective
Multi-modal Cardiac Magnetic Resonance
Diagnostic testA comprehensive "one-stop" scanning protocol using 3.0T MRI scanners. The protocol includes Cine imaging, T1/T2 mapping, Late Gadolinium Enhancement (LGE), first-pass perfusion, and 4D flow sequences to evaluate cardiac structure, function, myocardial tissue characteristics, and hemodynamics.
Diagnostic Efficacy of Multi-modal CMR Parameters (AUC)
The Area Under the Receiver Operating Characteristic Curve (AUC) will be calculated to assess the diagnostic performance of multi-modal CMR parameters (specifically T1 mapping and extracellular volume \[ECV\]) in differentiating various types of cardiovascular diseases (e.g., ischemic vs. non-ischemic). The final clinical diagnosis based on ESC/ACC guidelines will serve as the gold standard.
Correlation Between Imaging Parameters and Clinical Indicators
To evaluate the Pearson or Spearman correlation coefficients between multi-modal CMR parameters (such as LGE, T1/T2 mapping, 4D flow metrics) and clinical functional/laboratory indicators.
Incidence of Major Adverse Cardiovascular Events (MACE)
To evaluate the occurrence rate of MACE during the follow-up period. MACE is defined as a composite of cardiac death, readmission for heart failure, malignant arrhythmia (sustained ventricular tachycardia/ventricular fibrillation), and non-fatal myocardial infarction.
Model performance for virtual native enhancement across different sequence acquisition protocols
Model performance for virtual LGE generation across sequence protocols: Structural Similarity Index Measure (SSIM)
Model performance for virtual LGE generation across sequence protocols: Peak Signal-to-Noise Ratio (PSNR)
Peak Signal-to-Noise Ratio (PSNR) will be evaluated to compare virtual LGE algorithms trained on pre- vs. post-contrast Cine and 3-slice vs. 9-slice T1 mapping protocols.
Click on the lead sponsor to view all of their trials.
