Artificiella neurala nätverk underlättar svåra medicinska beslut. Exempel från pankreatologin
Sammanfattning
Artificial neural networks (ANNs) is a nonlinear pattern recognition technique inspired by the function of the human brain. ANNs have been used successfully in analysis of complex medical relationships, for example prognostic predictions of various forms of cancer. Acute pancreatitis and pancreatic cancer are commonly recognized to be extremely difficult conditions in making diagnostic and prognostic predictions with traditional methods. The two pancreatic disorders thus represent suitable candidates for analysis by ANNs. We hereby present a systematic review of the applications of ANNs in the treatment of acute pancreatitis and pancreatic cancer.