Skip to main content
. 2019 Oct 30;9(11):673. doi: 10.3390/biom9110673

Table 2.

Univariate analysis of clinicopathological features: deep learning classified cancer cases versus misclassified cancer cases. PSA = Prostate-specific antigen, TPV = Total prostate volume, PSAD = PSA density, WBC = White blood cell, Hb = Hemoglobin, Plt = Platelet, LDH = Lactate dehydrogenase, ALP = Alkaline phosphatase, Ca = Calcium, SD = Standard deviation.

Cancer Cases: N = 54 Classified Cases Misclassified Cases Univariate
(p Value)
Number of cases, (%) 92.6 7.4
Age, years, mean ± SD 67.4 ± 6.9 67.5 ± 7.3 0.96
PSA, ng/mL, mean ± SD 14.2 ± 11.9 21.6 ± 13.5 0.07
TPV, mL, mean ± SD 27.9 ± 10.7 23.0 ± 10.2 0.66
PSAD, ng/mL/cm3, mean ± SD 0.59 ± 0.64 1.17 ± 0.85 0.07
Gleason score, (%) 0.03
<8 60.0 0.0
≥8 40.0 100.0
Clinical stage, (%) 0.21
≤T2 80.0 50.0
≥T3 20.0 50.0
Pathological stage, (%) 0.63
≤T2 44.0 25.0
≥T3 56.0 75.0
WBC, 103/μL, mean ± SD 6074 ± 1248 5150 ± 656 0.12
Hb, g/dl, mean ± SD 14.5 ± 1.2 13.8 ± 0.7 0.08
Plt, 103/μL, mean ± SD 21.8 ± 5.0 18.3 ± 2.5 0.14
LDH, U/L, mean ± SD 180 ± 34.9 179 ± 45.4 0.93
ALP, U/L, mean ± SD 208 ± 56 249 ± 161 0.75
Ca, mg/dL, mean ± SD 9.3 ± 0.43 9.1 ± 0.26 0.29