Table 2 Comparison of clinical baseline data across training, internal testing, and external testing sets.

From: Multi-center study: ultrasound-based deep learning features for predicting Ki-67 expression in breast cancer

怂

Training set (Nā€‰=ā€‰557)

Internal test set (Nā€‰=ā€‰239)

External test set (Nā€‰=ā€‰133)

P-value

Age (years)

53.1 (10.3)

52.5 (10.3)

53.6 (9.75)

0.774

Height (cm)

156 (6.47)

156 (7.57)

156 (6.19)

0.983

Weight (kg)

56.4 (8.09)

56.5 (8.77)

56.1 (7.92)

0.981

BMI (kg/m2)

23.1 (3.38)

23.2 (3.86)

23.0 (3.25)

0.960

T staging

Ā 1

144 (25.9%)

64.0 (26.8%)

34.0 (25.6%)

0.978

Ā 2

314 (56.4%)

139 (58.2%)

75.0 (56.4%)

Ā 3

55.0 (9.9%)

20.0 (8.4%)

10.0 (7.5%)

Ā 4

44.0 (7.9%)

16.0 (6.7%)

14.0 (10.5%)

M staging

Ā 0

515 (92.5%)

228 (95.4%)

128 (96.2%)

0.251

Ā 1

42.0 (7.5%)

11.0 (4.6%)

5.00 (3.8%)

N staging

Ā 0

369 (66.2%)

141 (59.0%)

81.0 (60.9%)

0.062

Ā 1

129 (23.2%)

82.0 (34.3%)

33.0 (24.8%)

Ā 2

59.0 (10.6%)

16.0 (6.7%)

19.0 (14.3%)

Ki67

Ā Low

148 (26.6%)

81.0 (33.9%)

38.0 (28.6%)

0.223

Ā High

409 (73.4%)

158 (66.1%)

95.0 (71.4%)

EGFR

Ā Negative

425 (76.3%)

182 (76.2%)

104 (78.2%)

0.999

Ā Low

126 (22.6%)

54.0 (22.6%)

28.0 (21.1%)

Ā High

6.00 (1.1%)

3.00 (1.3%)

1.00 (0.8%)

ER

Ā Negative

259 (46.5%)

92.0 (38.5%)

63.0 (47.4%)

0.184

Ā Positive

298 (53.5%)

147 (61.5%)

70.0 (52.6%)

HER2

Ā Negative

181 (32.5%)

89.0 (37.2%)

48.0 (36.1%)

0.884

Ā Low

224 (40.2%)

91.0 (38.1%)

54.0 (40.6%)

Ā High

152 (27.3%)

59.0 (24.7%)

31.0 (23.3%)

PR

Ā Negative

226 (40.6%)

93.0 (38.9%)

57.0 (42.9%)

0.940

Ā Low

237 (42.5%)

104 (43.5%)

59.0 (44.4%)

Ā High

94.0 (16.9%)

42.0 (17.6%)

17.0 (12.8%)

Back to article page