An ellipsoidal shell model for volume estimation of the right ventricle from magnetic resonance images

RATIONALE AND OBJECTIVES: I developed a volume estimation technique for the crescentic volume of the right ventricle (RV) of the heart. A geometric model was desired to avoid the lengthy data collection and reduction required by Simpson's rule. METHODS: An ellipsoidal shell model was developed that requires only simple mathematics and that resembles the RV shape. RV cast volumes were obtained by water displacement and Simpson's rule and model-based calculations using magnetic resonance (MR) imaging. RESULTS: Model-based estimates correlated well with water displacement volumes (r = 0.924), with a slope not significantly different from unity. Simpson's rule results showed a higher correlation (r = 0.994), but it required longer acquisition and processing. Geometric irregularity in the RV shape required no modification in mathematics. CONCLUSION: The model provides reliable RV volume estimates from two MR image planes. The mathematics provides a simple approach to a relatively complex, crescentic shape. Short times for data acquisition and analysis suggest the potential for time savings during routine clinical measurements.