Estimation and Identifiability of Model Parameters in Human Nociceptive Processing Using Yes-No Detection Responses to Electrocutaneous Stimulation

Abstract

Healthy or pathological states of nociceptive subsystems determine different stimulus-response relations measured from quantitative sensory testing. In turn, stimulus-response measurements may be used to assess these states. In a recently developed computational model, six model parameters characterize activation of nerve endings and spinal neurons. However, both model nonlinearity and limited information in yes-no detection responses to electrocutaneous stimuli challenge to estimate model parameters. Here, we address the question whether and how one can overcome these difficulties for reliable parameter estimation. First, we fit the computational model to experimental stimulus-response pairs by maximizing the likelihood. To evaluate the balance between model fit and complexity, i.e., the number of model parameters, we evaluate the Bayesian Information Criterion. We find that the computational model is better than a conventional logistic model regarding the balance. Second, our theoretical analysis suggests to vary the pulse width among applied stimuli as a necessary condition to prevent structural non-identifiability. In addition, the numerically implemented profile likelihood approach reveals structural and practical non-identifiability. Our model-based approach with integration of psychophysical measurements can be useful for a reliable assessment of states of the nociceptive system.

Dynamic Modeling of Mitochondrial Membrane Potential Upon Exposure to Mitochondrial Inhibitors

“…We also analyzed structural model identifiability in order to study whether model parameters can be determined uniquely, by performing a profile likelihood analysis ( Yang et al, 2016 ; Raue et al, 2009 ) on the parameters γ and α ( Supplementary Text ). We found that the compound decay parameter γ is structurally identifiable ( Supplementary Figure S8 ), but the ion leakage rate α is not ( Supplementary Figure S9 ).…”

“…In a PL approach one calculates a confidence interval, yet it also helps to identify parameters that are structurally non-identifiable ( Raue et al, 2009 ; Yang et al, 2016 ). The confidence interval for a parameter is determined by computing the dependence of the maximum likelihood (i.e., the two-fold logL, or NPL ) on the parameter that is being profiled (i.e., fixed at different values).…”

Nonparametric tests for equality of psychometric functions

“…This was the motivation for the test developed by Logvinenko et al (2012) and it is a frequent concern in empirical studies that require the collection of large amounts of data across several sessions, where the observers' sensory state can vary across sessions (for empirical evidence to this effect see, e.g., García-Pérez, 2010;Leek, Hanna, & Marshall, 1991;von Dincklage, Olbrich, Baars, & Rehberg, 2013). A formal test is surely more dependable than judging by eye whether the shape described by data from different sessions look alike (e.g., Hutsell & Jacobs, 2013;Oliveira & Machado, 2008;Yang, Meijer, Buitenweg, & van Gils, 2016). In this type of application, each observer's data on each experimental condition are analyzed separately with I standing for the number of sessions at which data had been collected.…”