| Autor: |
Zdenko, Sonicki, Ante, Cvitković, Karen L, Edwards, Marica, Miletić-Medved, Dubravka, Cvorisćec, Vladimir, Babus, Bojan, Jelaković |
| Rok vydání: |
2009 |
| Předmět: |
|
| Zdroj: |
Studies in health technology and informatics. 150 |
| ISSN: |
0926-9630 |
| Popis: |
The aim of this paper was to assess relationship between possible endemic nephropathy (EN) markers visually by the CoPlot methodology, and to illustrate this promising data analysis approach. From 912 screened persons in 3 Croatian endemic villages, 25 persons were diagnosed as confirmed EN patients, 371 as non-EN, and the remainder were classified as suspected of having EN, or at risk. Data on 25 confirmed EN patients were matched with appropriate non-EN examinees. All records with missing data were excluded, resulting in 35 subjects with complete data on the 13 key EN variables for CoPlot mapping. CoPlot solution met the accepted goodness of fit measure thresholds. Result showed relationship between EN markers, identifying some nearly duplicated variables, and possible outliers needing some subsequent analysis. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
