Analysis of the applicability of artificial neural networks for studying blood plasma: determination of magnesium ion concentration as a case study
-
Alexandre Liparini
Abstract
Artificial neural networks are suggested for use in predicting metal ion concentration in human blood plasma. Simulated and available experimental data are used to train the artificial neural network. Particularly, using 850 simulated samples, the network predicted the magnesium-free ion concentration with an average error smaller than 1%. Clinical data recently reported for 20 patients were considered and the artificial neural network predicted the concentration of free magnesium ion with an average error of about 6%. Overall, the approach of using artificial neural networks as an alternative or complementary strategy to deal with the analysis of human blood plasma can be useful for clinical diagnostics, if there is sufficient data to train the artificial neural network.
References
1. Andriolo A, Moreira SR, Silva LA, de Carvalho AB, Vieira JG, Ghiringhello MT, et al. Cálcio ionizado no soro: estimativa do intervalo de referência e condições de coleta. J Bras Patol Med Lab 2004; 40: 85–9. 10.1590/S1676-24442004000200007Search in Google Scholar
2. Glusker JP, Katz AK, Bock CW. Metal ions in biological systems. Rigaku J 1999; 16: 8–16. Search in Google Scholar
3. May PM, Linder PW, Williams DR. Computer simulation of metal-ion equilibria in biofluids: models for the low-molecular-weight complex distribution of calcium(II), magnesium(II), manganese(II), iron(III), copper(II), zinc(II), and lead(II) ions in human blood plasma. J Chem Soc Dalton 1977; 6: 588–95. 10.1039/dt9770000588Search in Google Scholar
4. Perrin DD. Multiple equilibria in assemblages of metal ions and complexing species: a model for biological systems. Nature 1965; 206: 170–8. 10.1038/206170a0Search in Google Scholar
5. Perrin DD, Sayce IG. Computer calculation of equilibrium concentrations in mixtures of metal ions and complexing species. Talanta 1967; 14: 833–42. 10.1016/0039-9140(67)80105-XSearch in Google Scholar
6. Ingri N, Kakolowi W, Sillen LG, Warnqvis B. High-speed computers as a supplement to graphical method. V. HALTAFALL a general program for calculating composition of equilibrium mixtures. Talanta 1967; 14: 1261–86. 10.1016/0039-9140(67)80203-0Search in Google Scholar
7. Jagner D, Sahlin E, Renman L. Experimental and computational study of species formed during electrochemical stripping oxidation of copper in chloride media determination of copper (II) in the ng l −1 range by stripping potentiometry. Talanta 1995; 42: 1447–55. 10.1016/0039-9140(95)01593-ZSearch in Google Scholar
8. Zeevaart JR. Metal-ion speciation in blood plasma as a tool in predicting the in vivo behaviour of potential bone-seeking radiopharmaceutics. Dissertation. Universiteit van Pretoria, South Africa: Delft University Press, 2001:163 pp. Search in Google Scholar
9. Martin JD, Soria E, Camps G, Serrano AJ, Sepulveda JR, Jimenez V. Neural networks as effective techniques in clinical management of patients: some case studies. Trans Inst Measur Control 2004; 26: 169–83. 10.1191/0142331204tm118oaSearch in Google Scholar
10. Astion ML, Wilding P. Application of neural networks to the interpretation of laboratory data in cancer-diagnosis. Clin Chem 1992; 38: 34–8. 10.1093/clinchem/38.1.34Search in Google Scholar
11. Queraltó J, Torres J, Tura M, Hernandez C, Mangues M, Cardenete J. Plasma vancomycin concentrations predicted by neural networks. Clin Chem 2000; 46: A187–8. Search in Google Scholar
12. Reis MA, Sinisterra RD, Belchior JC. An alternative approach based on artificial neural networks to study controlled drug release. J Pharm Sci 2004; 93: 418–30. 10.1002/jps.10569Search in Google Scholar PubMed
13. Wang W, Bian ZZ, Yan LF, Su J. A novel individual blood glucose control model based on Mixture of Experts neural networks. Lect Notes Comput Sci 2004; 3174: 453–8. 10.1007/978-3-540-28648-6_72Search in Google Scholar
14. Queraltó JM, Torres J, Guinot M. Neural networks for the biochemical prediction of bone mass loss. Clin Chem Lab Med 1999; 37: 831–8. 10.1515/CCLM.1999.125Search in Google Scholar
15. Papik K, Molnar B, Fedorcsak P, Schaefer R, Lang F, Sreter L, et al. Automated prozone effect detection in ferritin homogeneous immunoassays using neural network classifiers. Clin Chem Lab Med 1999; 37: 471–6. 10.1515/CCLM.1999.076Search in Google Scholar
16. Tafeit E, Reibnegger G. Artificial neural networks in laboratory medicine and medical outcome prediction. Clin Chem Lab Med 1999; 37: 845–53. 10.1515/CCLM.1999.128Search in Google Scholar
17. Ichihara K, Sato K. Evidence-based laboratory interpretation system built on a large collection of case records with well-defined diagnoses. Clin Chem Lab Med 2001; 39: 1035–44. 10.1515/CCLM.2001.169Search in Google Scholar
18. Stephan C, Cammann H, Semjonow A, Diamandis EP, Wymenga LF, Lein M, et al. Multicenter evaluation of an artificial neural network to increase the prostate cancer detection rate and reduce unnecessary biopsies. Clin Chem 2002; 48: 1279–87. 10.1093/clinchem/48.8.1279Search in Google Scholar
19. Vertosick FT, Rehn T. Predicting behavior of an enzyme-linked immunoassay model by using commercially available neural-network software. Clin Chem 1993; 39: 2478–82. 10.1093/clinchem/39.12.2478Search in Google Scholar
20. Burtis CA. Converging technologies and their impact on the clinical laboratory. Clin Chem 1996; 42: 1735–49. 10.1093/clinchem/42.11.1735Search in Google Scholar
21. de Viterbo VD, Belchior JC. Artificial neural networks-applied for studying metallic complexes. J Comp Chem 2001; 22: 1691–701. 10.1002/jcc.1124Search in Google Scholar
22. de Viterbo VD, Belchior JC. Redes neurais artificiais aplicadas no estudo de sistemas aquáticos envolvendo íons e ligantes. In: 25 a Reunião Anual da Sociedade Brasileira de Química, 2002, Poços de Caldas. Livro de Resumos. São Paulo: Sociedade Brasileira de Qumica 002;1:AB114. Search in Google Scholar
23. Ni Y, Chen S, Kokot S. Spectrophotometric determination of metal ions in electroplatin solutions in the presence of EDTA with the aid of multivariate calibration and artificial neural networks. Anal Chim Acta 2002; 463: 305–16. 10.1016/S0003-2670(02)00437-3Search in Google Scholar
24. Djurhuus MS, Gram J, Petersen PH, Klitgaard NA, Bollerslev J, Beck-Nielsen H. Biological variation of serum and urinary magnesium in apparently healthy males. Scand J Clin Lab Invest 1995; 55: 449–58. 10.1080/00365519509075394Search in Google Scholar PubMed
25. Hristova EN, Cecco S, Niemela JE, Rehak NN, Elin RJ. Analyzer-dependent differences in results for ionized calcium, ionized magnesium, sodium, and pH. Clin Chem 1995; 41: 1649–53. 10.1093/clinchem/41.11.1649Search in Google Scholar
26. Saris NE, Mervaala E, Karppanen H, Khawaja JA, Lewenstam A. Magnesium. An update on physiological, clinical and analytical aspects. Clin Chim Acta 2000; 294: 1–26. 10.1016/S0009-8981(99)00258-2Search in Google Scholar
27. Willcocks JP, Mulquiney PJ, Ellory JC, Veech RL, Radda GK, Clarke K. Simultaneous determination of low free Mg 2+ and pH in human sickle cells using P-31 NMR spectroscopy. J Biol Chem 2002; 277: 49911–20. 10.1074/jbc.M207551200Search in Google Scholar
28. Huskens J, Main M, Malloy CR, Sherry AD. The determination of magnesium in human blood plasma by P-31 magnetic resonance spectroscopy using a macrocyclic reporter ligand. Biochim Biophys Acta 1997; 1336: 434–44. 10.1016/S0304-4165(97)00054-8Search in Google Scholar
29. Haykin S. Neural networks: a comprehensive foundation. New Jersey, USA: Prentice Hall, 1999:842 pp. Search in Google Scholar
30. Bevington PR. Data reduction and error for the physical sciences. New York, USA: McGraw Hill, 1969:336 pp. Search in Google Scholar
31. Press WH, Teukolsky SA, Veterling WT, Flannery BP. Numerical recipes in Fortran: the art of scientific computing. Cambridge, UK: Cambridge University Press, 1992:963 pp. Search in Google Scholar
32. Silvert W, Baptist M. Can neuronal networks be used in data-poor situations? In: Lek S, Guégan JF, editors. Artificial neuronal networks; application to ecology and evolution. Berlin: Springer-Verlag, 2000:262 pp. Search in Google Scholar
33. Reibnegger G, Wachter H. Self-organizing neural networks – an alternative way of cluster analysis in clinical chemistry. Clin Chim Acta 1996; 248: 91–8. 10.1016/0009-8981(95)06269-6Search in Google Scholar
34. Oliveira MF, Mendes DQ, Ferrari LI, Vasconcelos AT. Ribosome binding site recognition using neural networks. Genet Mol Biol 2004; 4: 644–50. 10.1590/S1415-47572004000400028Search in Google Scholar
35. Campos FF, Rolett JS. The exponents method for calculating equilibrium concentrations of complex species in solution. J Comp Chem 1995; 16: 534–44. 10.1002/jcc.540160503Search in Google Scholar
36. Ichikawa H. Hierarchy neural networks as applied to pharmaceutical problems. Adv Drug Deliv Rev 2003; 55: 1119–47. 10.1016/S0169-409X(03)00115-7Search in Google Scholar
©2005 by Walter de Gruyter Berlin New York
Articles in the same Issue
- The future of clinical laboratories: more testing or knowledge services?
- Apolipoprotein A5 and triglyceridemia. Focus on the effects of the common variants
- Genotyping of the 19-bp insertion/deletion polymorphism in the 5′ flank of β-hydroxylase gene by dissociation analysis of allele-specific PCR products
- APOE genotypes and dyslipidemias in a sample of the Portuguese population
- Serum proteins profile as an indicator of malignancy: multivariate logistic regression and ROC analyses
- Mean fluorescence intensity rate is a useful marker in the detection of paroxysmal nocturnal hemoglobinuria clones
- Alterations of glycosaminoglycan metabolism in the development of diabetic complications in relation to metabolic control
- Erythrocyte zinc content in critically ill patients
- Systemic levels of interleukin-6 and matrix metalloproteinase-9 in patients with multiple myeloma may be useful as prognostic indexes of bone disease
- Analysis of the applicability of artificial neural networks for studying blood plasma: determination of magnesium ion concentration as a case study
- The effect of long-term storage on measured plasma lactate concentrations and prospective lactate results from a multicenter trial of antiretroviral therapy
- Evaluation of renal function in intensive care: plasma cystatin C vs. creatinine and derived glomerular filtration rate estimates
- Multicentre physiological reference values for some urinary component-to-creatinine (creatininium) concentration ratios
- Diagnostic and analytical performance of a screening panel for allergy
- Preliminary evaluation of the Iris IQ™ 200 automated urine analyser
- The Asp298 but not the C-786 genotype of the endothelial nitric oxide synthase is reduced with age in healthy Swiss men
- Identification of preanalytical mistakes in the stat section of the clinical laboratory
- Congress of Clinical Chemistry and Laboratory Medicine 2nd Annual Conference of the German United Society for Clinical Chemistry and Laboratory Medicine (DGKL) together with the Austrian Society for Laboratory Medicine and Clinical Chemistry (ÖGLMKC) Jena, Germany, October 6–8, 2005
- SIBioC 2005 37th National Congress of the Italian Society of Clinical Biochemistry and Clinical Molecular Biology Rome, Italy, October 11–14, 2005
Articles in the same Issue
- The future of clinical laboratories: more testing or knowledge services?
- Apolipoprotein A5 and triglyceridemia. Focus on the effects of the common variants
- Genotyping of the 19-bp insertion/deletion polymorphism in the 5′ flank of β-hydroxylase gene by dissociation analysis of allele-specific PCR products
- APOE genotypes and dyslipidemias in a sample of the Portuguese population
- Serum proteins profile as an indicator of malignancy: multivariate logistic regression and ROC analyses
- Mean fluorescence intensity rate is a useful marker in the detection of paroxysmal nocturnal hemoglobinuria clones
- Alterations of glycosaminoglycan metabolism in the development of diabetic complications in relation to metabolic control
- Erythrocyte zinc content in critically ill patients
- Systemic levels of interleukin-6 and matrix metalloproteinase-9 in patients with multiple myeloma may be useful as prognostic indexes of bone disease
- Analysis of the applicability of artificial neural networks for studying blood plasma: determination of magnesium ion concentration as a case study
- The effect of long-term storage on measured plasma lactate concentrations and prospective lactate results from a multicenter trial of antiretroviral therapy
- Evaluation of renal function in intensive care: plasma cystatin C vs. creatinine and derived glomerular filtration rate estimates
- Multicentre physiological reference values for some urinary component-to-creatinine (creatininium) concentration ratios
- Diagnostic and analytical performance of a screening panel for allergy
- Preliminary evaluation of the Iris IQ™ 200 automated urine analyser
- The Asp298 but not the C-786 genotype of the endothelial nitric oxide synthase is reduced with age in healthy Swiss men
- Identification of preanalytical mistakes in the stat section of the clinical laboratory
- Congress of Clinical Chemistry and Laboratory Medicine 2nd Annual Conference of the German United Society for Clinical Chemistry and Laboratory Medicine (DGKL) together with the Austrian Society for Laboratory Medicine and Clinical Chemistry (ÖGLMKC) Jena, Germany, October 6–8, 2005
- SIBioC 2005 37th National Congress of the Italian Society of Clinical Biochemistry and Clinical Molecular Biology Rome, Italy, October 11–14, 2005
