Implementation of Proteomics Biomarkers in Nephrology: From Animal Models to Human Application?

Preclinical animal models are extensively used in nephrology. In this review, the utility of performing proteome analysis of kidney tissue or urine in such models and transfer of the results to human application has been assessed. Analysis of the literature identified 68 relevant publications. Pathway analysis of the reported proteins clearly indicated links with known biological processes in kidney disease providing validation of the observed changes in the preclinical models. However, although most studies focused on the identification of early markers of kidney disease or prediction of its progression, none of the identified makers has made it to substantial validation in the clinic or at least in human samples. Especially in renal disease where urine is an abundant source of biomarkers of diseases of the kidney and the urinary tract, it therefore appears that the focus should be on human material based discovery studies. In contrast, the most valid information of proteome analysis of preclinical models in nephrology for translation in human disease resides in studies focusing on drug evaluation, both efficacy for translation to the clinic and for mechanistic insight.     

胆固醇检测技术研究现状

人体胆固醇水平过高和心血管疾病的发生有着密切的联系。近年来,随着我国社会经济的发展,高胆固醇血症的发病率呈上升趋势。对胆固醇进行准确、及时的检测可以有效预防心血管疾病。本文介绍了各种胆固醇检测方法的原理,总结了胆固醇生物传感器的研究现状,并展望了该领域未来的发展趋势。     

Driving in an urban environment,the stress response and effects of exercise

Driving may be detrimental to health, with one hypothesis suggesting that driving may elicit an acute stress response and, with repeated exposures, may become a chronic stressor. The present study examined the stress response to driving and the effectiveness of a prior exercise bout in dampening this response. Twenty healthy adults performed three tasks: control, driving and exercise plus driving. Heart rate (HR), heart rate variability (HRV), blood pressure (BP) and cortisol were measured to quantify the acute stress response to each condition. Data indicated a stress response to driving: HR was elevated and HRV was reduced during the driving task compared with control. HR was elevated and HRV was reduced comparing the exercise plus driving with the driving condition. BP and cortisol were not different among conditions. The potential of interventions, such as exercise, to counter daily stressors should be evaluated to safeguard long-term health.

Practitioner Summary: this study confirms that driving induces a stress response, with the exercise intervention providing mixed results (an increase in cardiovascular measures and a decrease in cortisol measure trending significance). Given the known consequences of stress and evidence that exercise can mitigate acute stress, further evaluation of exercise interventions is recommended.     

基于改进深度信念网络的心血管疾病预测研究

定量分析和预测有助于降低心血管疾病发病风险。传统预测模型在多分类非线性复杂因素的条件下,预测准确率会降低。采用浅层神经网络的方法,能够解决该问题。然而因浅层神经网络初始参数的随机性,带来了预测结果方差增大的新问题。提出一种基于改进深度信念网络的心血管疾病预测模型,利用重构误差,自主确定网络深度,结合无监督训练和有监督调优,在提高模型预测准确率的同时保证稳定性。对UCI数据库中的Statlog (Heart)和Heart Disease Dataset数据集独立进行30次实验,结果显示预测准确率的均值分别为91.26%、89.78%,预测准确率的方差分别为5.78、4.46。     

基于改进深度神经网络的心血管疾病预测

心血管疾病是威胁人类健康的常见疾病,为了能够更加准确地对其预测,本文在传统DNN模型基础上进行优化改进,提出定向正则的深度神经网络(TR-DNN)模型,通过改进原有深度神经网络模型所存在的缺陷,使其能够更好地对心血管疾病数据集进行训练并测试,进一步实现心血管疾病预测任务。实验表明该模型在数据集训练上的表现良好,并且在测试集上取得优秀的结果。最后,将TR-DNN与SVM、RF、XGBoost模型在同一数据集进行结果比较,TR-DNN模型的各项评价指标均优于其它模型,在准确率方面相较传统DNN模型提高1.507个百分点,召回率提高1.57个百分点,特异度提高2.54个百分点,精确率提高1.51个百分点。因此,TR-DNN模型可以应用于心血管疾病的预测。     

基于 Asp ．net 的慢性肾脏病分期管理系统的实现与改进

近年来,慢性肾脏病（Chronic Kidney Disease ,CKD）的发病率呈逐年增长趋势,且患者对慢性肾脏病的知晓率非常低,以致病情恶化,错过了治疗最佳时机。论文设计并实现了基于 Asp ．net 的慢性肾脏病分期管理系统,从系统设计、系统架构、关键技术及改进、系统运行等方面介绍了实现过程。通过该系统的应用,在临床上能方便医生有效管理患者,并能掌握每个患者疾病的发展情况;同时,能够提高患者对病情的了解,增加患者的知晓率和防治慢性肾脏病的知识。系统对该种疾病的防治有重要的意义。