Table 1.

Characteristics and demographics of subjects used to train and validate the Videssa Breast model

Training setValidation setP
N469194
Age, median (range)58 (40–75)58 (40–75)0.94e
Race
 Caucasian41288%17088%0.27f
 Black/African American327%179%
 Asian71.5%52%
 American Indian/Alaska Native/Hawaiian/Pacific Islander71.5%10.5%
 Othera112%10.5%
BI-RADS Assessment
 315533%5930%0.35f
 430064%12565%
 5143%105%
Biopsied subjectsb3261330.37f
 BI-RADS 3(22)(5)
 BI-RADS 4(290)(118)
 BI-RADS 5(14)(10)
Benign breast conditions3841560.66f
 Procedure-confirmed benign(237)(93)
 Presumed benignc(143)(61)
 Lobular carcinoma in situd (LCIS)(4)(2)
Breast cancer (% Incidence)8518.1%3819.6%
 Invasive carcinoma (BC)(50)(30)
 DCIS(35)(8)
  • NOTE: Includes women over age 50 and women with high FSH.

  • aMulticultural or not reported.

  • bIncludes cyst aspiration and/or biopsy.

  • cPresumed all noncancer participants to be Benign.

  • dLCIS participants were categorized as noncancer (Benign).

  • eStatistical significance assessed by unpaired t test.

  • fStatistical significance assessed by Fisher exact test or χ2 (based on group size).