Prenatal diagnosis of cerebral lesions acquired in utero and with a late appearance

Although no precise figures are available, many congenital brain lesions arise from intrauterine disruption, frequently due to obstetric complications. The most common entities include intracranial hemorrhage, ischemic lesions, thrombosis of venous vessels and infections. Accurate prenatal diagnosis is possible in many of these cases. However, the findings may be subtle, particularly in the early stage of the disruptive process. Identification of these conditions requires therefore specific expertise, the combination of fetal neurosonography and magnetic resonance, and frequently there is a need for serial examinations. Targeted diagnostic imaging should be offered to obstetric patients with conditions predisposing to prenatal cerebral insults. Copyright © 2009 John Wiley & Sons, Ltd.     

Isolated sulfite oxidase deficiency: mutation analysis and DNA-based prenatal diagnosis

Isolated sulfite oxidase deficiency is an autosomal recessive, neurological disorder resulting from a defect in SUOX, the gene encoding the enzyme that catalyzes the terminal reaction in the sulfur amino acid degradation pathway. In its classical, severe form, sulfite oxidase deficiency leads to intractable seizures, severe and progressive brain pathology and death at an early age. We report here on clinical features and mutational analysis of the genetic defect in a newborn with sulfite oxidase deficiency. Cultured fibroblasts from this patient exhibited no detectable sulfite oxidase activity, and a unique four base pair deletion was present in the cDNA isolated from the same source. Identification of the same genetic defect in a heterozygous state in each of the parents and the monitoring of subsequent pregnancies in this family by DNA-based prenatal diagnosis are also described. The deletion mutation was identified in a homozygous state in uncultured chorionic villus tissue from the second pregnancy that was subsequently terminated. In the third pregnancy, the presence of sulfite oxidase activity and identification of the mutation in a heterozygous state suggested that the fetus was not affected. This pregnancy resulted in the birth of a normal child. Copyright © 2002 John Wiley & Sons, Ltd.     

基于人工免疫算法的离散隐马尔科夫故障诊断模型优化

目的解决离散隐马尔科夫模型在行星齿轮箱故障诊断中的自适应性和泛化性问题。方法建立人工免疫优化模型,将包含易被误判样本的多样本集作为抗原,以其正确识别率为适应度函数,不断对初始观测矩阵进行增殖、变异和选择,获得识别率最高时的初始观测矩阵,利用人工免疫算法对隐马尔科夫故障诊断模型的初始观测矩阵进行优化。结果将该方法应用于行星齿轮箱的故障诊断中,通过不同工况下的对比试验、单样本和多样本优化对比试验,验证了优化后的隐马尔科夫故障诊断模型的诊断效果。结论优化后的隐马尔科夫故障诊断模型具有更好的适应性,诊断精度显著提高。     

基于免疫机理的化工园区安全生产预警机制设计

为从预警角度预防化工园区安全生产事故,基于生物免疫机理,先应用仿生学方法从整体分析化工园区的"生理结构",再从局部分析化工园区安全生产事故预警体系与生物免疫系统的相似性,首次提出包括诊断识别子系统、预警信号子系统、协调控制子系统和记忆库的化工园区安全生产事故免疫预警系统;设计了一种新的包括预警指标参数输入、模型运算和结果输出3个功能模块的化工园区安全生产事故预警机制。预警机制的输入模块实现预警指标数据输入与处理的功能,运算模块实现预警模型运算的功能,输出模块给出量化评价和图形分析2种结果。基于此设计,将构建预警指标体系,建立化工园区安全生产现状预警和未来预测模型。