论文部分内容阅读
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 diagnostic system for antineoplastic drug adverse reactions (ADR) based on genetic neural network was proposed. Based on this, the myelosuppressive toxicity induced by LEVF regimen was investigated. All the relevant indexes were screened out and a good prediction was made. Solve the neural network structure selection and “over-fitting” problem. The system can be used for clinical diagnosis of ADR, ADR can also be used for drug evaluation.