Skip to main content
. 2021 Apr 1;23(4):e22394. doi: 10.2196/22394

Table 1.

Characteristics of 37 studies included in the systematic review.

Characteristics Studies, n Patients (average over the number of studies), n
Study type 
  Prospective 8 2210 (276.25)
  Retrospective 29 6414 (221.17)
Data set type
  Private data set 33 7760 (235.15)
  Public database (SPIE-AAPM-NCIa PROSTATEx challenge) 2 399 (199.5)
  Mixed (private and public) data set 2 465 (232.5)
Classification algorithms
  Random forest 4 1621(405.25)
  Regression-based models 20 4678 (233.9)
  Partial least squares discriminant analysis (PLS-DA) 2 180 (90)
  Linear discriminant analysis (LDA) 1 53
  Support vector machine (SVM) 2 65 (32.5)
  Classification and regression tree (CART) 1 67
  Artificial neural networks (ANNs) 2 1012 (506)
  Deep neural networks (DNNs) 1 195
  Convolutional neural networks (CNNs) 3 696 (232)
  Deep learning: SNCSAEb 1 57
Predictor type
  Multiparametric MRIc 20 5058 (252.9)
  Genetic or molecular biomarker 13 3132 (240.92)
    Urine 6 930 (155)
    Serum 3 901 (300.3)
    Semen 2 108 (54)
    Tissue 2 800 (400)
  Clinical data 4 2812 (703)
Validation method 
  Internal validation 29 6540 (225.52)
  External validation 3 1380 (460)
  Internal and external validation 1 364
  Unknown 5 704 (140.8)

aSPIE-AAPM-NCI: International Society for Optics and Photonics–American Association of Physicists in Medicine–National Cancer Institute.

bSNCSAE: stacked nonnegativity constraint sparse autoencoders.

cMRI: magnetic resonance imaging.