An artificial neural network for analysing the survival of patients with colorectal cancer.

摘要: An internet/web based artificial neural network has been developed for use by practicing clinical oncologists and medical researchers as part of a programme to aid decision making eventually, the management treatment individual patients with colorectal cancer. We have configured implemented Partial Likelihood Artificial Neural Network (PLANN) trained it predict cancer related survival in confirmed using database provided Clinical Resource Audit Group (CRAG) Scotland. The reliability PLANN was evaluated against Kaplan- Meier (KM) actual plots shows close agreement them. applied networks (ANNs) their associated analytical techniques healthcare, special reference suffering from common solid cancers. There is increasing complexity staging these cancers, requiring specialist, multidisciplinary knowledge, management. believe that systems such will become more readily available clinicians emergence web grid-secure technology, which potential link large scientific data sets various sources institutions. To date, ANNs varying types used, mainly research rather than routine oncology. Their usefulness investigated diagnosis, spread disease prognosis breast, ovarian, gastrointestinal, bronchial, prostatic ovarian cancers (1-3). In breast shown be significantly accurate predicting primary site (4). there no reported studies on formulate plans this remains long- term aim current interdisciplinary work our group Dundee physicists Manchester. So far, we ANN exposing existing one type (colorectal), where outcome included base known over 5-year follow-up period. This paper deals verification prediction web-based system.