Synthetic Haematocrit for Myocardial Extracellular Volume Mapping in Clinical Practice at 1.5 T and 3 T

Background/Objectives: Synthetic haematocrit (HCT) permits cardiovascular magnetic resonance (CMR) extracellular volume fraction (ECV) estimation without contemporaneous laboratory HCT. We evaluated four published equations in a clinical referral cohort. Methods: This retrospective audit included 67 adults scanned at 1.5 T (n = 34) or 3 T (n = 33), with laboratory HCT within 24 h. Two equations per field strength were assessed using correlation and Bland–Altman analysis with confidence intervals. Observer reproducibility of T1 and calculated synthetic ECV was assessed using intraclass correlation coefficients (ICCs). Post hoc exploratory analyses examined classification at ECV ≥ 30%, estimation-error associations and proportional bias. Results: Absolute mean ECV bias was below one percentage point for all equations, but model-specific limits of agreement extended from −4.47 to 4.19 percentage points. At the exploratory 30% threshold, classification agreement was 85.3% and 88.2% at 1.5 T and 75.8% for both equations at 3 T. Synthetic ECV intraobserver ICCs ranged from 0.853 to 0.949; interobserver ICCs were 0.936–0.944 at 1.5 T and 0.842–0.851 at 3 T. Shang HCT showed magnitude-dependent disagreement. Conclusions: Synthetic ECV may help assess myocardial tissue when timely laboratory haematocrit is unavailable, with low average differences from conventional ECV. However, the small groups limit precision and generalisability, and individual and observer differences prevent assuming interchangeability. Local validation is needed before using synthetic ECV for clinical decisions or monitoring changes over time.