Table 10 AD diagnostic performance obtained from every radiomic feature. The mean and standard deviation are obtained from fivefold cross-validation.

From: Enhancing foveal avascular zone analysis for Alzheimer’s diagnosis with AI segmentation and machine learning using multiple radiomic features

Feature

Fold 1

Fold 2

Fold 3

Fold 4

Fold 5

Mean (%)

Area

54

69

62

60

69

\(63.2\pm 5.0\)

Solidity

59

75

50

66

71

\(64.6\pm 8.0\)

Roundness

54

68

47

62

66

\(59.9\pm 7.0\)

Compactness

60

67

58

72

71

\(66.0\pm 5.0\)

Eccentricity

63

78

61

53

60

\(63.5\pm 8.0\)

All features

76

76

65

70

72

\(\bf{{72.2}\pm {4.2}}\)

Back to article page