A hyperbolic model for the diffusion of proteins through the blood-cerebrospinal fluid (CSF) barrier revolutionized clinical neurochemistry thirty years ago. The regression curves were informally parametrized based on physiologically-driven constraints. The current paper readdresses this issue with numerical optimization for unconstrained non-linear regression, implementing the Levenberg-Marquardt Algorithm (LMA). Astonishingly similar estimates are obtained, which reconfirms the concepts of H. Reiber proposed in 1990s. The LMA is discussed in the context of other optimization algorithms.
References
[1]
Reiber, H. (1994) Flow Rate of Cerebrospinal Fluid (CSF)—A Concept Common to Normal Blood-CSF Barrier Function and to Dysfunction in Neurological Diseases. JournaloftheNeurologicalSciences, 122, 189-203. https://doi.org/10.1016/0022-510x(94)90298-4
[2]
Crank, J. (1975) The Mathematics of Diffusion. 2nd Edition, Oxford University Press.
[3]
Reiber, H., Zeman, D., Kušnierová, P., Mundwiler, E. and Bernasconi, L. (2019) Diagnostic Relevance of Free Light Chains in Cerebrospinal Fluid—The Hyperbolic Reference Range for Reliable Data Interpretation in Quotient Diagrams. Clinica ChimicaActa, 497, 153-162. https://doi.org/10.1016/j.cca.2019.07.027
[4]
Levenberg, K. (1944) A Method for the Solution of Certain Non-Linear Problems in Least Squares. QuarterlyofAppliedMathematics, 2, 164-168. https://doi.org/10.1090/qam/10666
[5]
Marquardt, D.W. (1963) An Algorithm for Least-Squares Estimation of Nonlinear Parameters. JournaloftheSocietyforIndustrialandAppliedMathematics, 11, 431-441. https://doi.org/10.1137/0111030
[6]
Gavin, H. (2024) The Levenberg-Marquardt Algorithm for Nonlinear Least Squares Curve-Fitting Problems. https://people.duke.edu/~hpgavin/lm.pdf
[7]
Nocedal, J. and Wright, S.J. (2006) Numerical Optimization. 2nd Edition, Springer.
