Skip to main content

Table 1 Accuracy of proposed classifier for Alzheimer’s disease when conditioned on left and/or right hippocampi as well as area-based and/or conformal shape differences

From: Shape-based disease grading via functional maps and graph convolutional networks with application to Alzheimer’s disease

#

Right

Left

Both

\(h^a_{S,\omega }\)

\(h^c_{S,\omega }\)

\(h^{*}_{S,\omega }\)

\(h^a_{S,\omega }\)

\(h^c_{S,\omega }\)

\(h^{*}_{S,\omega }\)

\(h^a_{S,\omega }\)

\(h^c_{S,\omega }\)

\(h^{*}_{S,\omega }\)

120

77.9%

61.9%

79.2%

79.3%

67.3%

80.3%

80.4%

63.6%

81.5%

238

80.2%

70.2%

81.3%

80.2%

67.8%

81.9%

84.0%

71.4%

86.0%