Table 2 Baseline RGB characteristics in the NBI and WLI groups.

From: Deep learning diagnostics for bladder tumor identification and grade prediction using RGB method

 

Benigna

Low-and high-grade

p

Benigna, low-grade

High-grade

p

CNI

CIS

p

NBI

No

83 (6.5)

1198 (93.5)

 

630 (49.2)

651 (50.8)

 

361 (69.6)

158 (30.4)

 

Red

159.32 ± 25.32

158.99 ± 25.48

0.909

160.56 ± 24.95

157.51 ± 25.87

0.032

157.27 ± 24.92

157.75 ± 23.48

0.835

Green

110.20 ± 24.91

114.77 ± 23.93

0.094

114.88 ± 24.88

114.08 ± 23.16

0.550

115.05 ± 27.03

112.63 ± 21.59

0.278

Blue

84.63 ± 19.05

88.48 ± 18.94

0.074

88.58 ± 19.34

87.89 ± 18.61

0.513

88.50 ± 20.26

86.50 ± 16.15

0.231

WLI

No

198 (7.1)

2585 (92.9)

 

1261 (45.3)

1522 (54.7)

 

729 (74.0)

256 (26.0)

 

Red

180.35 ± 25.35

181.27 ± 23.66

0.598

177.09 ± 24.57

184.62 ± 22.55

< 0.001

179.04 ± 22.81

180.85 ± 20.28

0.262

Green

129.38 ± 23.28

130.75 ± 20.98

0.381

131.32 ± 22.50

130.10 ± 19.96

0.127

129.85 ± 19.29

131.73 ± 16.68

0.165

Blue

89.68 ± 32.86

94.45 ± 29.89

0.031

90.74 ± 31.53

96.91 ± 28.62

< 0.001

94.45 ± 27.34

96.64 ± 25.52

0.263

  1. Data are expressed as number (%), mean ± standard deviation.
  2. CIS carcinoma in situ, CNI chronic non-specific inflammation, NBI narrow-band image, RGB red/green/blue, WLI white-light image.
  3. aBenign defined cystitis cystica or cystitis grandularis except CNI.