基于遗传神经网络的抗肿瘤药物不良反应诊断系统

Diagnosing System for Adverse Drug Reactions of Anti- tumor Drugs Based on Genetic Neural Network

提出一种以遗传神经网络为基础的抗肿瘤药物不良反应(ADR)诊断系统,并以此考察了LEVF方案引起的骨髓抑制毒性,筛选了相关的各种指标,做出了较好的预测。解决了神经网络结构选择和“过度拟合”问题。该系统既可以用于临床的ADR诊断,也可以用于新药的ADR事件评价。

A diagnosing system for adverse drug reactions (ADR) of anti-tumor drugs based on genetic neural network is introduced . Bone marrow depression of LEVF treatment plan is studied on the system. Various related index are screened and superior predictions are acquired. Problems concerning network structural selection and excessive learning have been solved. The system can be used in clinical ADR diagnosis as well as ADR assessment for new drugs.