Quantitative Kinetic Models from Intravital Microcopy: A Case Study Using Hepatic Transport

Kenneth W. Dunn; Konstantinos Tsekouras; Meysam Tavakoli; Richard Day; Steve Presse

REVIEW 2 major objections 4 minor 14 references

Quantitative Kinetic Models from Intravital Microcopy: A Case Study Using Hepatic Transport

T0 review · 2 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Intravital microscopy fluorescence traces, modeled with compartment-specific measurement parameters and inferred by a generalized parameter-cascades scheme, can yield quantitative liver transport rates without invasive calibration.

desk verdict A legitimate adaptation of parameter cascades to hepatic IVM transport, with solid synthetic validation but a load-bearing insensitivity claim that is asserted, not demonstrated. read the letter →

arxiv 1908.10564 v1 pith:QABJ7HCY submitted 2019-08-28 physics.bio-ph q-bio.QM

Meysam Tavakoli , Konstantinos Tsekouras , Richard Day , Kenneth W. Dunn , Steve Presse This is my paper · ORCID

classification physics.bio-phq-bio.QM

keywords intravitalmicroscopyhepatictransportparametercascadesODEestimationmeasurementparametersfluoresceinglucuronidechronickidneydiseasetaurolithocholate

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that quantitative transport kinetics across liver compartments can be recovered from intravital microscopy fluorescence data alone. Because fluorescein emits differently in the sinusoid, hepatocyte, and canaliculus, and because both fluorescein and its glucuronide form contribute to the signal, the authors add compartment- and species-specific measurement parameters to a six-species ODE transport model and estimate those parameters together with the kinetic rates. The estimation tool is a generalized parameter-cascades procedure that avoids numerically solving the ODEs. Validation on simulated two-state, two-species, and a benchmark nerve-axon excitation model shows the rates are recoverable, and on real rat-liver data the method finds that 5/6 nephrectomy lowers sinusoid-to-hepatocyte uptake while taurolithocholate nearly abolishes hepatocyte-to-canaliculus efflux. If these estimates are right, live-imaging studies could quantify drug effects on specific hepatic transport steps without invasive sampling.

What carries the argument

The load-bearing device is a generalized parameter-cascades estimator. In parameter cascades, the ODE solution is approximated by a B-spline expansion whose coefficients (nuisance parameters) are fit in an inner optimization under a roughness penalty that enforces fidelity to the ODE, while the structural parameters—the kinetic rates—are updated in an outer optimization. The paper's generalization puts the entries of the measurement matrix $H$ into the structural parameter vector, so the per-compartment fluorescence visibility fractions ($\alpha$, $\beta$, $\gamma$ and their primed counterparts) are inferred from the same data instead of being calibrated separately. Identifiability analysis is used to decide which parameters must be prespecified: $\alpha = 0.5$, $\alpha' = 0.25$, and $k_{HT} = 0.5$ min$^{-1}$, with the remaining twelve parameters estimated.

What would settle it

Re-run the full hepatic-transport inference on the same sham, 5/6N, and TLC traces with $\alpha$, $\alpha'$, and $k_{HT}$ varied across their physiologically plausible ranges (for example $k_{HT}$ from 0.1 to 2 min$^{-1}$, and $\alpha$ and $\alpha'$ from 0.1 to 0.9). If the reported differences between sham and 5/6N uptake rates, or the near-zero canalicular rates in TLC, move by more than the quoted standard deviations, the central conclusion is not robust.

Watch

Extended reading notes

Core claim

The central claim is that the hidden species concentrations in each liver compartment and the measurement parameters that map them to observed fluorescence are jointly identifiable from noisy time traces, provided three parameters are fixed from outside knowledge. Implemented for hepatic transport, the model has six species—fluorescein and glucuronidated fluorescein in sinusoid, hepatocyte, and canaliculus—coupled by linear rates, with three fluorescence measurements described by a rectangular measurement matrix. On sham-control rat data the inferred sinusoid-to-hepatocyte uptake rate is $k_{S\to H} = 2.24$ min$^{-1}$; in 5/6 nephrectomy it drops to $2.07$ min$^{-1}$, while in taurolithocholate-treated rats the hepatocyte-to-canaliculus rate collapses to $k_{H\to C} = 0.0032$ min$^{-1}$ and canalicular loss to $k_C = 0.0026$ min$^{-1}$, matching the known cholestatic action of TLC. The paper also demonstrates on synthetic data that the estimation scheme recovers ground-truth parameters even when measurement noise is added.

Load-bearing premise

The load-bearing premise is that the three prespecified values taken from earlier studies—the two sinusoid visibility fractions ($\alpha=0.5$, $\alpha'=0.25$) and the hepatocyte glucuronidation rate ($k_{HT}=0.5$ min$^{-1}$)—are accurate enough that the inferred rates are insensitive to their exact values, an assertion the paper states without reporting a sensitivity analysis.

Editorial extensions

If this is right

Transport-step-specific rate constants can be assigned to disease or drug effects from image time series, not just from isolated-cell or vesicle assays.
Because measurement parameters are estimated, the method should work on data with different optical conditions or quencher environments, and on other fluorescent probes that change emission by compartment.
The near-zero canalicular rates recovered under taurolithocholate give a quantitative, image-based signature of cholestasis that could be tracked over time or across doses.
For chronic kidney disease, the specific reduction in sinusoid-to-hepatocyte uptake offers a concrete kinetic explanation for impaired hepatic clearance previously reported in 5/6N rats.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A natural extension, not carried out in the paper, is to replace the fixed $\alpha$, $\alpha'$, and $k_{HT}$ values with per-animal measurements from fluorescence lifetime imaging; this would turn the sensitivity question into a direct comparison rather than an assumption.
If the measurement matrix is identifiable from data, the scheme could generalize to spectrally distinct probes or to transporters with multiple substrates, using the same latent-species structure.
The model's linear kinetics could be replaced by saturable transport kinetics to capture transporter saturation; a testable prediction is that high-dose fluorescein should slow apparent uptake rates when uptake is saturable.
Comparing transporter-knockout animals with the same inference pipeline would test whether each transporter's removal shifts the corresponding edge rate constant, a prediction the current data does not resolve.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Colleague,

Quick take: this is a credible applied inference paper, not a methods breakthrough. The new piece is modest but real: they include compartment-specific measurement parameters as structural parameters in a parameter-cascades fit, and they push the framework through a six-species, three-observation hepatic transport model with three experimental conditions. The synthetic validation (two-state, generalized two-state, FitzHugh-Nagumo) is the strongest part of the paper. It shows the machinery recovers known ground truth at several noise levels, with error bars that grow sensibly as noise increases. That is real evidence the core estimator works.

The experimental conclusions are biologically plausible and align with prior physiology: 5/6N nephrectomy lowers sinusoid-to-hepatocyte uptake modestly, and TLC nearly abolishes canalicular efflux. The comparison of sham vs. 5/6N rates (kS→H 2.24 vs. 2.07 min−1) is not huge relative to the reported standard deviations, so I would not oversell that difference. The TLC effect on kH→C (0.0032 min−1) is stark and clearly consistent with the known mechanism.

Soft spots, in proportion:

1. The printed objective functions in Eqs. (4)–(5) omit the measurement matrix H. As written they compare yi to x̂i, but y is length r (here 3) and x̂ is length m (here 6). That is dimensionally inconsistent unless the hat implicitly means Hx̂. This is a typo-level fix, but in a methods paper it matters, and it is the kind of thing a referee should catch.

2. The insensitivity claim is load-bearing and unsupported. They prespecify α=0.5, α′=0.25, kHT=0.5 min−1 and say the conclusions are insensitive, but no sensitivity analysis appears in the main text or in what I can see of the SI. Because the sinusoid measurement is a weighted sum αS+α′S′, and kHT feeds the H′ source term, these choices can trade against the inferred rates. The stress-test note is right: this is the central quantitative vulnerability. It may well be that the rates are robust in practice, but the paper needs to show a profile or a small table demonstrating that before I trust Tables 4–6.

3. Minor: confidence intervals and identifiability details are deferred to SI. That is fine for a JPC B paper, but it does mean the main text's precision claims are not self-contained.

Overall, the method is honestly presented, the synthetic work is reproducible in spirit, and the biology is not overinterpreted. The missing sensitivity analysis is addressable and should be the main referee request. I would take this paper seriously and engage with it, but I would not cite the absolute rate values from the real-data tables until the sensitivity question is settled.

Recommendation: send to peer review. The synthetic validation alone justifies that. Ask for the H fix, a sensitivity table for α, α′, kHT, and a more explicit statement about how the reported standard deviations were computed. After that, it is publishable.

Best, [You]

Referee Report

2 major / 4 minor

Summary. The paper develops a parameter-cascade method to infer kinetic and measurement parameters of ordinary differential equation models for hepatic fluorescein transport from intravital microscopy (IVM) fluorescence time traces. The model couples unmodified and glucuronidated fluorescein species across three compartments (sinusoid, hepatocyte, canaliculus), with measurement parameters mapping hidden species concentrations to observed fluorescence and accounting for compartment-specific quenching. The method is validated on synthetic data from a two-state model, a generalized two-state model with two species, and the FitzHugh-Nagumo model, with increasing noise levels. The authors then apply the method to previously published IVM data from sham, 5/6-nephrectomized (5/6N), and taurolithocholate (TLC)-treated rats, reporting reduced sinusoid-to-hepatocyte uptake in 5/6N and near-zero canalicular efflux in TLC. Identifiability requires prespecifying α=0.5, α′=0.25, and kHT=0.5 min−1, and the paper asserts without demonstrating that the results are insensitive to these choices.

Significance. If the method holds, it offers a route to quantitative in vivo transport rates from fluorescence data alone, avoiding absolute concentration calibration. The inclusion of measurement parameters in a parameter-cascade framework is a genuine methodological extension, and the synthetic validation on multiple nonlinear systems supports the inference machinery. The application to disease models addresses a physiologically important problem. However, the central quantitative claims for the real data rest on three externally fixed parameters, and the insensitivity assertion is not backed by a sensitivity analysis. Additionally, the printed objective functions omit the measurement matrix, making the method as stated internally inconsistent. With these points addressed, the paper would be a useful contribution to quantitative intravital microscopy and ODE parameter estimation.

major comments (2)

[Materials and Methods, Eqs. (4)-(5)] The objective functions used in the inner and outer optimization loops are written as sums of squared differences between the data y_i and the spline approximation xhat_i of the state vector. This is inconsistent with the measurement model in Eq. (2), y = Hx + ε, because y_i are fluorescence measurements in three compartments while x_i are six hidden species concentrations. As written, minimizing |y_i − xhat_i|^2 would only be correct if H were the identity, which contradicts the entire purpose of estimating measurement parameters. The equations must be revised to include H, e.g., |y_i − H xhat_i|^2, and the dimensions of H clarified (Eq. (2) says H is m×n but Eq. (15) shows it is 3×6). Without this correction, the method as described is not reproducible and the statistical meaning of the outer criterion is unclear.
[Results, Tables 4-6 and the paragraph on prespecified parameters] The identifiability analysis (cited to SI Appendix B) forces the authors to fix α=0.5, α′=0.25, and kHT=0.5 min−1. The text states that 'our quantitative conclusions are insensitive to exact parameter estimates used initially for kHT, α and α′' but gives no sensitivity analysis, profile likelihood, or supporting table. This is load-bearing: the sinusoid measurement is yS = αS + α′S′, so changing α/α′ directly reallocates the observed sinusoid signal between unmodified and glucuronidated fluorescein; similarly, kHT enters the source term for H′ and can trade off against k′S→H and k′H→S. If the reported differences between sham, 5/6N, and TLC (e.g., kS→H 2.24 vs 2.07 min−1, kH→C 0.0032 min−1) are sensitive to these fixed inputs, the biological conclusions would be compromised. The authors should either provide a systematic sensitivity analysis over a physiologically plausible range of these parameters and show that the inferred rates and their confidence intervals are stable, or explicitly state the dependence and weaken the conclusions.

minor comments (4)

[Eq. (14)] In the differential equation for dH′/dt, the term representing loss to the canaliculus is written as k_H→C H′ rather than k′_H→C H′. This is inconsistent with the model schematic and with the dC′/dt equation, which uses k′_H→C. Please correct this apparent typo.
[General typos] The paper contains several typographical errors, including 'Microcopy' in the title, 'reducig' in the Results text, 'mehods' in the Discussion, and 'These coe cients' in the duplicated passage near the end. A careful proofreading pass is needed.
[Algorithm 1 and Eq. (4)] In Algorithm 1, the inner optimization is written as minimizing Jin(c′|θ′, λ) using a prime on c that is not defined elsewhere; Eq. (4) also uses c without saying whether it is the vector of all spline coefficients. Please clarify the notation.
[Standard deviations in Tables 4-6] The experimental tables report standard deviations for the estimated parameters, but the main text does not explain how these are computed. The SI is cited for confidence intervals, but a brief statement in the main text about the procedure (e.g., bootstrap or asymptotic covariance) would improve interpretability.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the hepatic transport rates are genuine outputs of the cascade optimization, not restatements of the prespecified identifiability inputs.

full rationale

The derivation chain is self-contained with respect to circularity. The method is validated on synthetic data with known ground truth (Tables 1–3), and the real-data inference uses external IVM data from Ref. 29. The parameters α=0.5, α′=0.25, and kHT=0.5 min−1 are fixed for identifiability, not estimated from the data whose conclusions they would otherwise predetermine; the reported rates are outputs of the cascade optimization, not restatements of these inputs. The paper's assertion that conclusions are insensitive to these values is not supported by a sensitivity analysis, but that is a robustness gap rather than a circular reduction. Self-citations (Refs. 51, 52, 61, 62, 77) are contextual or methodological and do not carry the central argument. No equation in the paper reduces a predicted quantity to a fitted or cited input by construction.

Assumptions & free parameters 11 free parameters · 5 assumptions · 0 invented entities

The central inference rests on a compartmental ODE model with a linear measurement map (six species to three fluorescence channels), on prespecified values for three unidentifiable parameters, and on the assumed adequacy of spline approximations. No new physical entities are introduced; the hidden species are known metabolites of fluorescein.

free parameters (11)

kS→H (sinusoid-to-hepatocyte uptake rate) = 2.2440 (sham), 2.0661 (5/6N), 1.6657 (TLC) min^-1
Fitted to IVM fluorescence time traces; central to the claim that 5/6N slows hepatic uptake.
kH→S (hepatocyte-to-sinusoid efflux rate) = 2.1501 (sham), 2.2067 (5/6N), 0.9180 (TLC) min^-1
Fitted; part of the back-flow branch.
kH→C (hepatocyte-to-canaliculus secretion rate) = 0.1726 (sham), 0.1580 (5/6N), 0.0032 (TLC) min^-1
Fitted; TLC condition intended to block this rate.
kC (canaliculus loss rate) = 0.3802 (sham), 0.2753 (5/6N), 0.0026 (TLC) min^-1
Fitted; near zero under TLC, interpreted as canalicular block.
k'S→H, k'H→S, k'H→C, k'C (glucuronidated-species rates) = Tables 4-6 (sham, 5/6N, TLC) min^-1
Fitted; the primed species are glucuronidated fluorescein, transported by the same scheme with distinct rates.
β, γ, β', γ' (measurement parameters for hepatocyte and canaliculus) = β=0.4731, γ=0.1393, β'=0.1340, γ'=0.3763 (sham); other cases in Tables 5-6
Fitted as structural parameters; represent fraction of species fluorescence observed per compartment.
α (sinusoid measurement parameter) = 0.5 (prespecified)
Preset to resolve structural unidentifiability, chosen from other physiological experiments.
α' (sinusoid measurement parameter, glucuronidated) = 0.25 (prespecified)
Preset for identifiability.
kHT (fluorescein glucuronidation rate in hepatocyte) = 0.5 min^-1 (prespecified)
Preset for identifiability; authors claim results insensitive to this value but do not demonstrate.
noise variance σ² = not stated
Assumed known in a pre-calibration step per Eq. (2); value not reported.
regularization parameter λ = not stated
Controls smoothing penalty in J_in; no selection criterion given.

assumptions (5)

domain assumption The liver can be represented by three compartments (sinusoid, hepatocyte, canaliculus) with no direct sinusoid-to-canaliculus transport.
Stated in Results when constructing the hepatic transport model; if violated, estimated rates absorb additional pathways.
domain assumption Transport kinetics are linear first-order ODEs with constant rate coefficients and a single glucuronidation step in the hepatocyte.
Equations (13)-(14); saturation (Michaelis-Menten) is not included despite being mentioned in the discussion.
domain assumption Measured fluorescence is a linear function of hidden species concentrations with constant per-compartment proportionality factors (measurement parameters) plus white noise of known variance.
Equation (2) and Eq. (15); environmental quenching and dye forms are folded into the constant factors.
ad hoc to paper Prespecifying α, α′, and kHT resolves structural unidentifiability and leaves the remaining 12 parameters identifiable from three fluorescence time traces.
Results section: 'To resolve model unidentifiability, we pre-specify kHT as well as the measurement parameters α and α′'; the identifiability analysis is deferred to SI and the insensitivity claim is not shown.
standard math B-spline expansions with penalized least squares provide an adequate approximation to the ODE solutions for the chosen λ.
Eqs. (3)-(4); standard functional-data-analysis result from Ramsay et al. (2007) used here without modification.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Quantitative Kinetic Models from Intravital Microcopy: A Case Study Using Hepatic Transport." pith.science (2026). https://pith.science/paper/QABJ7HCY

@misc{pith2026190810564,
  author       = {Pith},
  title        = {Pith review of: Quantitative Kinetic Models from Intravital Microcopy: A Case Study Using Hepatic Transport},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QABJ7HCY}},
  note         = {Machine review of arXiv:1908.10564}
}

read the original abstract

The liver performs critical physiological functions, including metabolizing and removing substances, such as toxins and drugs, from the bloodstream. Hepatotoxicity itself is intimately linked to abnormal hepatic transport and hepatotoxicity remains the primary reason drugs in development fail and approved drugs are withdrawn from the market. For this reason, we propose to analyze, across liver compartments, the transport kinetics of fluorescein-a fluorescent marker used as a proxy for drug molecules-using intravital microscopy data. To resolve the transport kinetics quantitatively from fluorescence data, we account for the effect that different liver compartments (with different chemical properties) have on fluorescein's emission rate. To do so, we develop ordinary differential equation transport models from the data where the kinetics are related to the observable fluorescence levels by "measurement parameters" that vary across different liver compartments. On account of the steep non-linearities in the kinetics and stochasticity inherent to the model, we infer kinetic and measurement parameters by generalizing the method of parameter cascades. For this application, the method of parameter cascades ensures fast and precise parameter estimates from noisy time traces.

Figures

Figures reproduced from arXiv: 1908.10564 by the authors.

**Figure 2.** Fits for the Synthetic Model Shown in Fig. (1). For each compartment the results are shown after adding white noise with variance 0.05 (A), 0.1 (B) and 0.2 (C). The dots are generated data points, the dash lines are theoretical curves, and the solid lines are the fits on our data. AU here stands for arbitrary units. 2- Generalized Two States System: We continue testing our approach by generalizing the previous examp… view at source ↗

**Figure 3.** Fits for Generalized Two States Systems. For each compartment the results of adding white noise with variance 0.05 (A), 0.1 (B) and 0.2 (C) have been shown. The dots are generated data points, dash lines are theoretical curves, and the solid lines are the fits on our data. V , and a recovery variable R summarizing outward currents. In this case we setup simulated data for parameter values [a, b, c] = [0.2, 0.2, 3] a… view at source ↗

Figures from the paper (3 more)

**Figure 4.** Figure 4: Fit to the FitzHugh-Nagumo Model Supplemented by a Measurement Model. As before, we consider the following different noise variance levels: 0.05 (A), 0.1 (B) and 0.2 (C). Details in text. RESULTS Full Hepatic Transport Model We now construct a model of hepatic transpor…

**Figure 5.** Figure 5: Schematic of Hepatic Transport Model Used. We consider five transport rates for fluorescein: kS→H for the transport from sinusoid to the hepatocyte, kH→S from the hepatocyte back to the sinusoid, kH→C from the hepatocyte to the canaliculus, and a loss rate kC from the …

**Figure 6.** Figure 6: Results of Studies of Fluorescein Transport in Rat Model Sham Control, 5/6N, and Taurolithocholate (TLC) Treated Livers. In (A) we have sample images from a set of collected images from the liver of a rat as the sham control (without removing the kidneys) during intrav…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

14 extracted references · 14 canonical work pages

[1]

M.; Buer, J.; Dumoutier, L.; Renauld, J.-C.; Stockinger, B

(1) Veldhoen, M.; Hirota, K.; Westendorf, A. M.; Buer, J.; Dumoutier, L.; Renauld, J.-C.; Stockinger, B. The Aryl Hydrocarbon Receptor Links TH17-Cell-Mediated Autoim- munity to Environmental Toxins. Nature 2008, 453, 106–109. 24 (2) Brent, J. A.; Rumack, B. H. Role of Free Radicals in Toxic Hepatic Injury II. Are Free Radicals the Cause of Toxin-Induced ...

work page 2008
[2]

5 Perfusion After 1 minute After 6 minutes 0 2 4 6 8 10 120 500 1000 1500 2000 2500 3000 3500 Time(min) Signal (AU) 0 2 4 6 8 10 120 500 1000 1500 2000 2500 3000 3500Signal (AU) 0 2 4 6 8 10 120 500 1000 1500 2000 2500 3000 3500Signal (AU) 38

work page 2000
[8]

Extended Kalman Filtering of a Special Class of Diﬀerential-algebraic Equation Systems. Ind. Eng. Chem. Res. 2016, 56, 206–215. 31 (61) Jazani, S.; Sgouralis, I.; Press´ e, S. A Method for Single Molecule Tracking Using a Conventional Single-focus Confocal Setup. J. Chem. Phys. 2019, 150, 114108. (62) Jazani, S.; Sgouralis, I.; Shafraz, O. M.; Sivasankar,...

work page 2016
[74]

(98) Cao, J.; Ramsay, J. O. Generalized Proﬁling Estimation for Global and Adaptive Penalized Spline Smoothing. Comput. Stat. Data Anal. 2009, 53, 2550–2562. (99) Hines, K. E. A Primer on Bayesian Inference for Biophysical Systems. Biophys. J. 2015, 108, 2103–2113. 35 (100) Hines, K. E.; Middendorf, T. R.; Aldrich, R. W. Determination of Parameter Iden- t...

work page 2009
[120]

G.; Pe˜ nalva, G

(80) Roma, M. G.; Pe˜ nalva, G. L.; Ag¨ uero, R. M.; Garay, E. A. R. Hepatic Transport of Organic Anions in Taurolithocholate-induced Cholestasis in Rats. J. Hepatol. 1994, 20, 603–610. 33 (81) Petzinger, E. Transport of Organic Anions in the Liver. An Update on Bile Acid, Fatty Acid, Monocarboxylate, Anionic Amino Acid, Cholephilic Organic Anion, and Ani...

work page 1994
[228]

(49) Calderhead, B.; Girolami, M.; Lawrence, N. D. Accelerating Bayesian Inference Over Nonlinear Diﬀerential Equations with Gaussian Processes. Advances in neural infor- mation processing systems. 2009; pp 217–224. (50) Gupta, S.; Kihara, Y.; Maurya, M. R.; Norris, P. C.; Dennis, E. A.; Subrama- niam, S. Computational Modeling of Competitive Metabolism B...

work page 2009
[339]

Impulses and Physiological States in Theoretical Models of Nerve Mem- brane

(70) FitzHugh, R. Impulses and Physiological States in Theoretical Models of Nerve Mem- brane. Biophys. J. 1961, 1, 445–466. 32 (71) Nagumo, J.; Arimoto, S.; Yoshizawa, S. An Active Pulse Transmission Line Simulating Nerve Axon. IEEE. Proc. IRE 1962, 50, 2061–2070. (72) Cao, J.; Wang, L.; Xu, J. Robust Estimation for Ordinary Diﬀerential Equation Mod- els...

work page 1961
[1267]

These coe cients are estimated with penalized smoothing splines with a roughness penalty term

37 Graphical TOC Entry Perfusion After 1 minute After 6 minutes 0 2 4 6 8 10 120 500 1000 1500 2000 2500 3000 3500 Time(min) Signal (AU) 0 2 4 6 8 10 120 500 1000 1500 2000 2500 3000 3500Signal (AU) 0 2 4 6 8 10 120 500 1000 1500 2000 2500 3000 3500Signal (AU) ,53 works for linear or nonlinear dynamics, straightforwardly estimates measurement param- eters...

work page 2000

Show all 14 references

[1929]

Riemann Manifold Langevin and Hamiltonian Monte Carlo Methods

(47) Girolami, M.; Calderhead, B. Riemann Manifold Langevin and Hamiltonian Monte Carlo Methods. J. R. Stat. Soc. Series B Stat. Methodol. 2011, 73, 123–214. (48) Dondelinger, F.; Husmeier, D.; Rogers, S.; Filippone, M. ODE Parameter Inference using Adaptive Gradient Matching ...

2011
[1974]

Nonlinear Parameter Estimation: A Case Study Comparison

(40) Biegler, L.; Damiano, J.; Blau, G. Nonlinear Parameter Estimation: A Case Study Comparison. AIChE J. 1986, 32, 29–45. (41) Ramsay, J. O. Functional data analysis. Encyclopedia of Statistical Sciences 2004, 4, 37–111. (42) Brunel, N. J. Parameter Estimation of ODEs via Non...

1986 arXiv
[1997]

(69) Bellman, R.; ˚Astr¨ om, K. J. On Structural Identiﬁability.Math. Biosci. 1970, 7, 329–

1970
[1999]

D.; Callaghan, J

(17) Bjornsson, T. D.; Callaghan, J. T.; Einolf, H. J.; Fischer, V.; Gan, L.; Grimm, S.; Kao, J.; King, S. P.; Miwa, G.; Ni, L. The Conduct of in Vitro and In Vivo Drug- Drug Interaction Studies: A Pharmaceutical Research and Manufacturers of America (PhRMA) Perspective. Drug ...

2003
[2000]

O.; Hooker, G.; Campbell, D.; Cao, J

(38) Ramsay, J. O.; Hooker, G.; Campbell, D.; Cao, J. Parameter Estimation for Diﬀeren- tial Equations: A Generalized Smoothing Approach. J. R. Stat. Soc. Series B Stat. Methodol. 2007, 69, 741–796. (39) Bard, Y. Nonlinear Parameter Estimation ; Academic Press: New York, NY, USA,

2007
[2008]

G.; Brown, M

(25) Presson, R. G.; Brown, M. B.; Fisher, A. J.; Sandoval, R. M.; Dunn, K. W.; Lorenz, K. S.; Delp, E. J.; Salama, P.; Molitoris, B. A.; Petrache, I. Two-photon Imaging within the Murine Thorax without Respiratory and Cardiac Motion Artifact. Am. J. Pathol. 2011, 179, 75–82. ...

2011

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

M.; Buer, J.; Dumoutier, L.; Renauld, J.-C.; Stockinger, B

(1) Veldhoen, M.; Hirota, K.; Westendorf, A. M.; Buer, J.; Dumoutier, L.; Renauld, J.-C.; Stockinger, B. The Aryl Hydrocarbon Receptor Links TH17-Cell-Mediated Autoim- munity to Environmental Toxins. Nature 2008, 453, 106–109. 24 (2) Brent, J. A.; Rumack, B. H. Role of Free Radicals in Toxic Hepatic Injury II. Are Free Radicals the Cause of Toxin-Induced ...

work page 2008

[2] [2]

5 Perfusion After 1 minute After 6 minutes 0 2 4 6 8 10 120 500 1000 1500 2000 2500 3000 3500 Time(min) Signal (AU) 0 2 4 6 8 10 120 500 1000 1500 2000 2500 3000 3500Signal (AU) 0 2 4 6 8 10 120 500 1000 1500 2000 2500 3000 3500Signal (AU) 38

work page 2000

[3] [8]

Extended Kalman Filtering of a Special Class of Diﬀerential-algebraic Equation Systems. Ind. Eng. Chem. Res. 2016, 56, 206–215. 31 (61) Jazani, S.; Sgouralis, I.; Press´ e, S. A Method for Single Molecule Tracking Using a Conventional Single-focus Confocal Setup. J. Chem. Phys. 2019, 150, 114108. (62) Jazani, S.; Sgouralis, I.; Shafraz, O. M.; Sivasankar,...

work page 2016

[4] [74]

(98) Cao, J.; Ramsay, J. O. Generalized Proﬁling Estimation for Global and Adaptive Penalized Spline Smoothing. Comput. Stat. Data Anal. 2009, 53, 2550–2562. (99) Hines, K. E. A Primer on Bayesian Inference for Biophysical Systems. Biophys. J. 2015, 108, 2103–2113. 35 (100) Hines, K. E.; Middendorf, T. R.; Aldrich, R. W. Determination of Parameter Iden- t...

work page 2009

[5] [120]

G.; Pe˜ nalva, G

(80) Roma, M. G.; Pe˜ nalva, G. L.; Ag¨ uero, R. M.; Garay, E. A. R. Hepatic Transport of Organic Anions in Taurolithocholate-induced Cholestasis in Rats. J. Hepatol. 1994, 20, 603–610. 33 (81) Petzinger, E. Transport of Organic Anions in the Liver. An Update on Bile Acid, Fatty Acid, Monocarboxylate, Anionic Amino Acid, Cholephilic Organic Anion, and Ani...

work page 1994

[6] [228]

(49) Calderhead, B.; Girolami, M.; Lawrence, N. D. Accelerating Bayesian Inference Over Nonlinear Diﬀerential Equations with Gaussian Processes. Advances in neural infor- mation processing systems. 2009; pp 217–224. (50) Gupta, S.; Kihara, Y.; Maurya, M. R.; Norris, P. C.; Dennis, E. A.; Subrama- niam, S. Computational Modeling of Competitive Metabolism B...

work page 2009

[7] [339]

Impulses and Physiological States in Theoretical Models of Nerve Mem- brane

(70) FitzHugh, R. Impulses and Physiological States in Theoretical Models of Nerve Mem- brane. Biophys. J. 1961, 1, 445–466. 32 (71) Nagumo, J.; Arimoto, S.; Yoshizawa, S. An Active Pulse Transmission Line Simulating Nerve Axon. IEEE. Proc. IRE 1962, 50, 2061–2070. (72) Cao, J.; Wang, L.; Xu, J. Robust Estimation for Ordinary Diﬀerential Equation Mod- els...

work page 1961

[8] [1267]

These coe cients are estimated with penalized smoothing splines with a roughness penalty term

37 Graphical TOC Entry Perfusion After 1 minute After 6 minutes 0 2 4 6 8 10 120 500 1000 1500 2000 2500 3000 3500 Time(min) Signal (AU) 0 2 4 6 8 10 120 500 1000 1500 2000 2500 3000 3500Signal (AU) 0 2 4 6 8 10 120 500 1000 1500 2000 2500 3000 3500Signal (AU) ,53 works for linear or nonlinear dynamics, straightforwardly estimates measurement param- eters...

work page 2000