Table 3 Summary of automated detection of cerebral malaria from retinal imaging

Li

FA

Weakly-supervised DL

LFL/PL

0.91 ± 0.02 ^a

0.97 ± 0.02 ^a

0.96 ± 0.02 ^a

0.94 ± 0.02 ^a

0.85 ± 0.02 ^a

Kurup

Colour

Transfer DL

Haemorrhage & Whitening

0.90

1

0.81

0.98

–

Chen

Colour

Supervised DL

Haemorrhage

–

–

–

0.94

0.995

Yan

FA

Superpixel-based TCV

LFL / PL

0.95

0.95

–

0.95

–

Zhao

FA

Superpixel-based TCV

LFL / PL

0.95

0.95

–

0.95

–

Rochim

FA

Saliency-based TCV

LFL / PL

-

0.9

0.98

–

–

Joshi

Colour

TCV

Haemorrhage

0.79

0.93

–

0.88

–

Whitening

0.67

0.77

–

0.75

–

MR

0.68

0.92

–

0.82

–

Zhao

FA

Saliency-based TCV

LFL / PL

0.93 ± 0.03

0.96 ± 0.02

0.91 ± 0.03

0.94 ± 0.02

0.82 ± 0.03

Zhao

FA

Saliency-based TCV

LFL / PL

0.83 ± 0.08

0.83 ± 0.03

0.81 ± 0.04

0.83 ± 0.05

–

MacCormick

FA

Spatial statistical model TCV

CNP

–

–

–

–

–

Joshi

Colour

Unsupervised TCV

Haemorrhage

0.73

0.96

–

0.89

–

Whitening

0.65

0.94

–

0.81

–

Vessel Discolouration

0.62

1

–

0.85

–

MR

0.95

1

–

0.97

–

Akram

Colour

TCV

Whitening

0.90

0.92

–

0.9

–

Zhao

FA

Saliency-based TCV

IVFD

0.75

0.74

0.74

0.74

–

Zhao

FA

Saliency-based TCV

LFL

0.95

–

–

–

–

PL

0.82

–

–

–

–

Vessel Leak

0.81

0.82

0.8

0.84

–

Joshi

Colour

TCV

Whitening

0.88 ^b

0.65 ^b

–

0.8

–

0.82 ^c

0.89 ^c

–

–

–

Ashraf

Colour

TCV

Whitening

–

–

0.92

–

–

Ashraf

Colour

TCV

Haemorrhage

0.95

0.96

0.97

0.95

–

Whitening

–

–

0.92

–

–

CWS

0.82

–

–

–

–

Agurto

Colour

TCV

Vessel Discolouration

0.94 ^b

0.67 ^c

0.85

0.87

–

0.68 ^b

1 ^c

0.85

–

–

Zheng

FA

Texture-based TCV

CNP

0.73 ± 0.14

0.91 ± 0.06

0.89 ± 0.04

–

–

Saleem

Colour

TCV

Haemorrhage

0.95

0.96

0.97

0.95

–

Joshi

Colour

Splat-based TCV

Haemorrhage

0.81

0.97

–

0.91

–