用电位多传感器系统测定尿离子成分。
Determination of urine ionic composition with potentiometric multisensor system.
机构信息
Chemistry Department, St. Petersburg State University, Universitetskaya nab. 7/9, Mendeleev Center, 199034 St. Petersburg, Russia; Bioanalytical Laboratory CSU "Analytical Spectrometry", St. Petersburg State Polytechnical University, Box 27, Gzhatskaya Street 27, 198220 St. Petersburg, Russia; ITMO University, Kronverkskiy pr., 49, 197101 St. Peterssburg, Russia.
Chemistry Department, St. Petersburg State University, Universitetskaya nab. 7/9, Mendeleev Center, 199034 St. Petersburg, Russia; ITMO University, Kronverkskiy pr., 49, 197101 St. Peterssburg, Russia.
出版信息
Talanta. 2015 Jan;131:556-61. doi: 10.1016/j.talanta.2014.08.030. Epub 2014 Aug 20.
The ionic composition of urine is a good indicator of patient's general condition and allows for diagnostics of certain medical problems such as e.g., urolithiasis. Due to environmental factors and malnutrition the number of registered urinary tract cases continuously increases. Most of the methods currently used for urine analysis are expensive, quite laborious and require skilled personnel. The present work deals with feasibility study of potentiometric multisensor system of 18 ion-selective and cross-sensitive sensors as an analytical tool for determination of urine ionic composition. In total 136 samples from patients of Urolithiasis Laboratory and healthy people were analyzed by the multisensor system as well as by capillary electrophoresis as a reference method. Various chemometric approaches were implemented to relate the data from electrochemical measurements with the reference data. Logistic regression (LR) was applied for classification of samples into healthy and unhealthy producing reasonable misclassification rates. Projection on Latent Structures (PLS) regression was applied for quantitative analysis of ionic composition from potentiometric data. Mean relative errors of simultaneous prediction of sodium, potassium, ammonium, calcium, magnesium, chloride, sulfate, phosphate, urate and creatinine from multisensor system response were in the range 3-13% for independent test sets. This shows a good promise for development of a fast and inexpensive alternative method for urine analysis.
尿液的离子组成是患者一般状况的良好指标，并可用于诊断某些医学问题，如尿路结石。由于环境因素和营养不良，登记的尿路病例数量不断增加。目前用于尿液分析的大多数方法都很昂贵，非常繁琐，并且需要熟练的人员。本工作研究了离子选择性和交叉敏感传感器的 18 个多传感器系统作为分析工具用于确定尿液离子组成的可行性。总共对来自尿路结石实验室的 136 个患者样本和健康人样本进行了多传感器系统以及毛细管电泳（作为参考方法）分析。实现了各种化学计量学方法将电化学测量数据与参考数据相关联。逻辑回归（LR）用于将样本分为健康和不健康两类，产生了合理的错误分类率。偏最小二乘回归（PLS）用于从电位数据定量分析离子组成。从多传感器系统响应中同时预测钠离子、钾离子、铵离子、钙离子、镁离子、氯离子、硫酸根离子、磷酸根离子、尿酸和肌酐的平均相对误差在独立测试集的范围内为 3-13%。这表明开发一种快速且廉价的尿液分析替代方法具有良好的前景。
Zotero 插件