Table 3 Cross-validated prediction error on the healthy reference set for three different models: Campens’ Z-score which uses BSA as a feature, the GP which uses BSA as a feature, and the GP that uses height and weight instead of BSA. Results are reported separately for two anatomical levels: SoV and AA.

From: Towards a more reliable assessment of aortic diameters using a Bayesian Z-score

Model

SoV

AA

MAE [mm]

RMSE [mm]

MAE [mm]

RMSE [mm]

Campens (BSA)

2.15

2.75

2.16

2.81

Bayesian GP (BSA)

2.14

2.75

2.15

2.80

Bayesian GP (Height + Weight)

2.13

2.73

2.14

2.79

  1. MAE = mean absolute error; RMSE = root-mean-square error (both in mm).
Back to article page