Skip to main content

Table 3 Confidence intervals of model performance

From: Enhancing magnetic resonance imaging-driven Alzheimer’s disease classification performance using generative adversarial learning

  1.5 T 3 T* 3 T* − 1.5 T
(a)
ADNI test [0.8968, 0.9172] [0.9304, 0.9336] [0.0203, 0.0297]
AIBL [0.9258, 0.9422] [0.9373, 0.9427] [0.0021, 0.0097]
NACC [0.8590, 0.8810] [0.9058, 0.9082] [0.0314, 0.0412]
(b)
ADNI test [0.9022, 0.9218] [0.9324, 0.9356] [0.0175, 0.0265]
AIBL [0.7347, 0.7833] [0.7467, 0.7593] [−0.0170, 0.0051]
NACC [0.8268, 0.8512] [0.8768, 0.8792] [0.0334, 0.0443]
  1. (a) 95% confidence intervals of the SS curves for 1.5-T-based model, 3-T*-based model, and difference of AUCs between 3-T* and 1.5-T-based models. (b) 95% confidence intervals of PR curves for 1.5-T-based model, 3-T*-based model and difference of AUCs between 3-T* and 1.5-T-based models