Extrapolation of Clinical Data from an Oral Glucose Tolerance Test Using a Support Vector Machine

Authors: Jianyin Lu, Masayoshi Seike, Wei Liu, Peihong Wu, Lihua Wang, Yihua Wu, Yasuhiro Naito, Hiromu Nakajima, Yasuhiro Kouchi

Abstract:

To extract the important physiological factors related to diabetes from an oral glucose tolerance test (OGTT) by mathematical modeling, highly informative but convenient protocols are required. Current models require a large number of samples and extended period of testing, which is not practical for daily use. The purpose of this study is to make model assessments possible even from a reduced number of samples taken over a relatively short period. For this purpose, test values were extrapolated using a support vector machine. A good correlation was found between reference and extrapolated values in evaluated 741 OGTTs. This result indicates that a reduction in the number of clinical test is possible through a computational approach.

Keywords: SVM regression, OGTT, diabetes, mathematical model

Digital Object Identifier (DOI): doi.org/10.5281/zenodo.1057087

References:

[1] Y. Naito, H. Ohno, A. Sano, H. Nakajima, and M. Tomita, "Construction of a simulation model of diabetes for pathophysiological analysis using E-Cell System," Genome Informatics, vol. 13, pp. 478-479, Dec. 2002.
[2] K .G. M. M. Alberti and P. Z. Zimmet, "Definition, diagnosis and classification of diabetes mellitus and its complications. part. 1: diagnosis and classification of diabetes mellitus. Provisional report of a WHO Consultation," Diabet Med, vol. 15, no. 7, pp. 539-553, Jul. 1998.
[3] C. Dalla Man, M. Campioni, K. S. Polonsky, R. Basu, R. A. Rizza, G. Toffolo, and C. Cobelli, "Two-hour seven-sample oral glucose tolerance test and meal protocol: minimal model assessment of beta-cell responsivity and insulin sensitivity in nondiabetic individuals," Diabetes, vol. 54, no. 11, pp. 3265-3273, Nov. 2005.
[4] C. Dalla Man, A. Caumo, and C. Cobelli, "The oral glucose minimal model: estimation of insulin sensitivity from a meal test," IEEE Trans Biomed Eng, vol. 49, no.5, pp. 419-429, May. 2002.
[5] G. Toffolo, E. Breda, M. K. Cavaghan, D. A. Ehrmann, K. S. Polonsky, and C. Cobelli, "Quantitative indexes of beta-cell function during graded up&down glucose infusion from C-peptide minimal models," Am J Physiol Endocrinol Metab, vol. 280, no. 1, pp. E2-E10, Jan. 2001.
[6] V. Vapnik, S. Golowich, and A. J. Smola, "A support vector method for function approximation, regression estimation, and signal processing," in Advances in Neural Information Processing Systems: Proceedings of the 1996 conference on Neural Information Processing Systems, vol. 9, M. C. Mozer, M. I. Jordan, and T. Petsche, Ed. Cambridge: MIT Press, 1997, pp. 281-287.
[7] ISO15197, "In vitro diagnostic test systems: Requirements for bloodglucose monitoring systems for self-testing in managing diabetes mellitus," May. 2003.
[8] C. Dalla Man, M. Camilleri, and C. Cobelli, "A system model of oral glucose absorption: validation on gold standard data," IEEE Trans Biomed Eng, vol. 53, no. 12 Pt1, pp. 2472-2478, Dec. 2006.