Comparison of two different artificial neural networks for prostate biopsy indication in two different patient populations.

@article{Stephan2007ComparisonOT,
  title={Comparison of two different artificial neural networks for prostate biopsy indication in two different patient populations.},
  author={Carsten Stephan and Chuanliang Xu and Patrik Finne and Henning Cammann and H H Meyer and Michael Lein and Klaus Jung and U E Stenman},
  journal={Urology},
  year={2007},
  volume={70 3},
  pages={596-601}
}
OBJECTIVES Different artificial neural networks (ANNs) using total prostate-specific antigen (PSA) and percentage of free PSA (%fPSA) have been introduced to enhance the specificity of prostate cancer detection. The applicability of independently trained ANN and logistic regression (LR) models to different populations regarding the composition (screening versus referred) and different PSA assays has not yet been tested. METHODS Two ANN and LR models using PSA (range 4 to 10 ng/mL), %fPSA… CONTINUE READING