Table 1 Demographics in cancer cases.

From: Applying Data-driven Imaging Biomarker in Mammography for Breast Cancer Screening: Preliminary Study

 

Train (n = 3101)

Validation (n = 619)

Test (n = 619)

P value

Density

   

0.7843

   almost entire fat

196 (6.32)

32 (5.2)

31 (5.0)

 

   scattered fibroglandular densities

640 (20.6)

136 (22.0)

137 (22.1)

 

   heterogeneous dense

1555 (50.2)

312 (50.4)

312 (50.4)

 

   extremely dense

710 (22.9)

139 (22.4)

139 (22.5)

 

Age

   

0.9941

   ≥50

1759 (56.7)

350 (56.5)

350 (56.5)

 

   <50

1342 (43.3)

269 (43.5)

269 (43.5)

 

Manufacturer

   

0.9351

   GE

1226 (39.5)

238 (38.5)

251 (40.6)

 

   Hologic

1032 (33.2)

200 (32.3)

198 (32.0)

 

   Siemens

843 (27.2)

181 (29.2)

170 (27.4)

 

Feature

    

   mass

1688 (54.4)

339 (54.8)

339 (54.8)

0.9806

   non mass

1413 (45.6)

280 (45.2)

280 (45.2)

 

   calcifications

1402 (45.2)

280 (45.2)

280 (45.2)

0.9999

   non calcifications

1699 (54.8)

339 (54.8)

339 (54.8)

 

Type

   

0.2767

   Invasive

2673 (86.2)

542 (87.56)

547 (88.37)

 

   Noninvasive

428 (13.8)

77 (12.44)

72 (11.63)

 

Size (invasive)

   

0.8409

   Size ≥20

1216 (45.5)

254 (46.9)

251 (45.9)

 

   Size <20

1457 (54.5)

288 (53.1)

296 (54.1)