Clinical problems, computational solutions
Richard F. Levine · Cancer · 2001
Computational approaches have become necessary to interpret the multiple parameters that may now be used for clinical decision making in the treatment of patients with cancer, including TNM classification, cancer markers, genes, and other blood tests. Artificial neural networks may be somewhat better than logistic regression techniques to deal with nonlinear and contingent relationships as well as the more chaotic data becoming available in computerized medical records systems.