Abstract

Background
Transthoracic echocardiography (TTE) is the standard method for the evaluation of the severity of aortic stenosis (AS). Valve effective orifice area (EOA) measured by the continuity equation is one of the most frequently used stenotic indices. However, TTE measurement of aortic valve EOA is not feasible or not reliable in a significant proportion of patients. Cardiovascular magnetic resonance (CMR) has emerged as a non-invasive alternative to evaluate EOA using velocity measurements. The objectives of this study were: 1) to validate a new CMR method using jet shear layer detection (JSLD) based on acoustical source term (AST) concept to estimate the valve EOA; 2) to introduce a simplified JSLD method not requiring vorticity field derivation.

Methods and results
We performed an in vitro study where EOA was measured by CMR in 4 fixed stenoses (EOA = 0.48, 1.00, 1.38 and 2.11 cm2) under the same steady flow conditions (4-20 L/min). The in vivo study included eight (8) healthy subjects and 37 patients with mild to severe AS (0.72 cm2 ≤ EOA ≤ 1.71 cm2). All subjects underwent TTE and CMR examinations. EOA was determinated by TTE with the use of continuity equation method (TTECONT). For CMR estimation of EOA, we used 3 methods: 1) Continuity equation (CMRCONT); 2) Shear layer detection (CMRJSLD), which was computed from the velocity field of a single CMR velocity profile at the peak systolic phase; 3) Single plane velocity truncation (CMRSPVT), which is a simplified version of CMRJSLD method. There was a good agreement between the EOAs obtained in vitro by the different CMR methods and the EOA predicted from the potential flow theory. In the in vivo study, there was good correlation and concordance between the EOA measured by the TTECONT method versus those measured by each of the CMR methods: CMRCONT (r = 0.88), CMRJSLD (r = 0.93) and CMRSPVT (r = 0.93). The intra- and inter- observer variability of EOA measurements was 5 ± 5% and 9 ± 5% for TTECONT, 2 ± 1% and 7 ± 5% for CMRCONT, 7 ± 5% and 8 ± 7% for CMRJSLD, 1 ± 2% and 3 ± 2% for CMRSPVT. When repeating image acquisition, reproducibility of measurements was 10 ± 8% and 12 ± 5% for TTECONT, 9 ± 9% and 8 ± 8% for CMRCONT, 6 ± 5% and 7 ± 4% for CMRJSLD and 3 ± 2% and 2 ± 2% for CMRSPVT.

Conclusion
There was an excellent agreement between the EOA estimated by the CMRJSLD or CMRSPVT methods and: 1) the theoretical EOA in vitro, and 2) the TTECONT EOA in vivo. The CMRSPVT method was superior to the TTE and other CMR methods in terms of measurement variability. The novel CMR-based methods proposed in this study may be helpful to corroborate stenosis severity in patients for whom Doppler-echocardiography exam is inconclusive.