Mitral valve repair is commonly used to address degenerative or functional changes to the mitral valve apparatus and surrounding ventricular anatomy. Preoperative transoesophageal echocardiogram (TOE) is routinely used to evaluate and identify the precise anatomic location of mitral valve pathology in order to guide repair. However, surgeons currently lack specific guidance regarding the approximate dimensions of the mitral valve they should aim for in order to achieve optimal valve function and avoid adverse outcomes. Therefore, through an observational study, we aimed to develop and test the accuracy of a preliminary mathematical model which represents the geometric relationship between various clinically relevant components of the mitral valve and its surrounding structures.

Using established trigonometric principles, the geometric relationship shared between several mitral valve components was represented in a two-dimensional (2D) model and described in a mathematical equation. The output variable of the model is the anteroposterior diameter of the mitral valve. To assess the accuracy of the mathematical model, we compared the model-predicted anteroposterior (AP) diameter against AP diameter measured by postoperative TOE in 42 cases.

The root mean squared error (RMSE) of model predicted AP diameter compared to measured AP diameter was 0.43 cm. The mean absolute percentage error (MAPE) of the model was 17.7%. In 34 out of 42 cases, model-predicted AP diameters were within 25% of AP diameters measured by postoperative TOE.

Preliminary testing of a simple mathematical model has shown its relative accuracy in representing the geometric relationship between several mitral valve variables. Further research and refinement of the model is required in order to improve its accuracy. We are encouraged that, with further improvement, the model has the potential for clinical application.