1Deutsches Herzzentrum der Charite (DHZC) Klinik für Angeborene Herzfehler - Kinderkardiologie Berlin, Deutschland; 2Technische Universität Berlin Institut für Mathematik, MA 8-3, TRR 109 Discretization in Geometry and Dynamics Berlin, Deutschland; 3Deutsche Herzzentrum der Charite (DHZC) Institut für kardiovaskuläre computer-assistierte Medizin Berlin, Deutschland
Background:
Assessment of right ventricular (RV) volume during the follow-up of patients with congenital heart defects (CHD) is crucial. Due to its complex geometry, transthoracic echocardiographic RV volumetry might remain challenging and MRI is widely used to assess RV volume. However, MRI does not provide bedside live-imaging, is time consuming, expensive and in patients with implanted devices often not applicable. Therefore, we aimed to establish a reliable, easy and fast to obtain method to calculate RV volume using two-dimensional transthoracic echocardiography (2D echo) and validate it with MRI data.
Methods:
A retrospective single-center trial was conducted at Deutsches Herzzentrum der Charité. Standard apical four-chamber and parasternal short-axis view was obtained in CHD patients during routine follow-up using 2D echo. RV volumes were calculated using (α+1)/(α+2)*SS'/d formula (S = RV area obtained in standard four-chamber view; S`= RV area obtained in standard short-axis view; d = Diameter of tricuspid valve anulus) and compared to RV volumes obtained by MRI.
Results:
22 patients (12 female) with various types of heart defects were included. As published before, RV volume quantification by 2D echo traditionally relies on the 2/3*S'*a formula, elucidated by an ellipsoidal shell model. However, in our cohort acquired results indicated a need to augment this formula, particularly in diastolic scenarios, with an additional constant factor for reliable data. This constant factor, unaccounted for in the model, necessitates an alternative approach. Therefore, we propose a model based on a rotational body of the curve xα. Accordingly, we calculated a constant correction factor α for systolic volume (αS = 3.1) and for diastolic volume (αD = 3.9), respectively. The associated formula (α+1)/(α+2)*SS'/d demonstrates superior performance in our experiments with a mean difference to MRI systolic volume of 4.84 ml (13.3%; R2=0,88) and a mean difference to MRI diastolic volume of 16.55 ml (19.2%; R2=0.88)
Conclusion:
The proposed method to determine RV volume using two-dimensional echocardiography allows distinct modeling in systolic and diastolic phases with excellent correlation to RV volume obtained by MRI. Thus, this promising approach provides a feasible, exact and fast method to assess RV volume on bedside, facilitating prompt and precise clinical decision-making.