Model, evidence & numerical checks
A transparent computational experiment in how nerve repair unfolds over space and time. Representative axon trajectories interact with a changing tissue environment, while structural arrival, pathway identity and maturation remain distinct.
The equations form an original reduced model informed by the studies below. They do not reproduce a published regeneration dataset and have not been validated for clinical, animal-treatment or device-efficacy predictions.
1. What the model represents
Local biological environment
The disconnected distal axon population loses integrity, becomes debris and is cleared. Aggregate macrophage activity and a Schwann repair-state variable interact with debris removal. The bridge gains prescribed vascular and Schwann support from its ends. Fibrosis supplies an adjustable obstruction term.
These normalized states preserve biological distinctions and causal hypotheses. They are not measurements of cell numbers, oxygen, cytokines, c-Jun expression or collagen concentration.
Individual axon trajectories
Each representative parent axon receives seeded traits for initiation, potential speed, caliber and pathway geometry. Its tip advances along a synthetic curve at a rate set by the local environment. Tortuosity converts arc-length growth into axial advance. The same seed and parameters reproduce the same experiment.
Paths are generated from geometry and alignment assumptions before integration. Emergent growth-cone steering, chemotaxis, cell adhesion forces, collateral sprouting and molecular target recognition are not solved.
Keep the endpoints separate
| Readout | Exact interpretation | What it cannot establish |
|---|---|---|
| Distal entry | A parent axon has crossed the distal repair boundary. | Contact with an appropriate end organ. |
| Target-distance arrival | The tip has reached the axial target coordinate: gap plus specified distal distance. | A functional neuromuscular junction or sensory ending. |
| Fascicle-matched arrival | The arrived axon’s destination fascicle index matches its original synthetic fascicle. | Recovery of its original motor unit, receptor, organ or motor/sensory branch. |
| Myelin maturity | A local time-since-arrival exponential; the headline mean is evaluated at distal entry among entered parents. | Measured sheath thickness, internodal length, g-ratio or conduction velocity. |
| Conduction surrogate | Target-arrival maturation multiplied by a seeded caliber factor, averaged over all original parents. | Action potentials, CMAP, force, sensation or recovery probability. |
| Target receptivity | An independent adjustable denervation-age diagnostic. | Target survival, a clinical deadline or a contact-rescue model. It does not gate growth. |
Motor and sensory labels identify representative fibers without encoding target-specific guidance. Noninitiating and stalled trajectories stay in the denominator and are not described as dead neurons.
2. Injury and repair architectures
| Scenario | Implemented distinction | Interpretation |
|---|---|---|
| Crush / preserved tracks | No gap or imposed repair delay; preserved fascicle alignment; distal degeneration and regrowth still occur. | Idealized preserved-pathway axonotmesis, not every crush severity. |
| Direct apposition | Cut ends touch with zero gap; alignment remains adjustable. | No surgical tension, suturing trauma or fascicular coaptation calculation. |
| Hollow conduit | No initial internal Schwann scaffold; prescribed ingrowth fronts populate the gap. | A generic architecture, without a commercial material or universal critical gap length. |
| Structured bridge | Internal support after placement, at an assumed architectural support fraction. | Generic supportive scaffold; not a fitted autograft, allograft, cellular or biohybrid treatment effect. |
| Delayed repair | Axotomy competence, recipient distal support and target clocks run from injury. The fresh bridge appears at placement and has its own support age. | Compare the same architecture at different delays with a common seed. |
| Unrepaired transection | Local sprouting without credited distal entry. | No spontaneous gap closure or neuroma morphology is resolved. |
Presets combine geometry with example coefficients. Comparing them does not estimate treatment efficacy. Pin a run and vary one parameter when investigating a specific causal assumption.
3. Equations and units
Time is days since injury. Axial and transverse coordinates are millimeters. Environmental states and maturation indices are dimensionless. Numerical coefficients below are explicit assumptions of this release.
Local reaction states
Q: disconnected distal axon integrity; D: debris; M: macrophage activity; S: support available to still-advancing axons; V: vascular support; F: fibrosis burden. Every state is dimensionless.
Assumption boundaryThe fields are local reaction variables with prescribed spatial bridge fronts. There is no diffusion equation, cell census, molecular concentration, or reinnervation feedback into S.
Fragmentation and debris removal
The fragmentation rate k rises smoothly from zero at day 1 to 0.7/day at day 2, then stays constant. Intact distal material transfers into debris, while baseline, macrophage, and Schwann activity clear it. The cleared fraction is tracked by mass bookkeeping.
Assumption boundaryRates are model assumptions. Electrical response disappearance in human nerves does not measure the same quantity as Q or D.
Aggregate macrophage activity
W = exp(−t/6) exp(−|x − g/2|/max(2, g/2)) is an injury-centered wound pulse; g is gap length in mm and t is days since injury.
Assumption boundaryThe state is not a macrophage count or immune-subtype model. Recruitment and resolution rates are illustrative coefficients.
Support, vascular, and fibrosis relaxation
τV is 3.5 days within the bridge and 1.5 days in the distal stump. S*, V*, and F* are prescribed local target states, not measured equilibrium values.
Assumption boundaryThis closure gives a responsive, bounded environment. It does not solve oxygen delivery, vessel growth, collagen deposition, or c-Jun kinetics.
Activation and chronic support loss
In the distal stump, S* = S0 A(t) Cs(t). Hs is the user-selected repair-support half-life. Bridge S* is also multiplied by the prescribed bridge occupancy. A newly inserted structured bridge uses its own time since repair for A and Cs; the distal recipient and host-derived conduit support retain time since injury.
Assumption boundaryThe 28-day plateau and half-life are assumptions, including in human-scale scenarios. Mouse and human molecular findings do not calibrate this curve.
Bridge occupancy from both stumps
tr is time after repair; d = min(x, g − x). After repair V* = V0(0.06 + 0.94BV), and S* = S0 A Cs [a + (1 − a)BS], with a = 0.72 for the structured-bridge scenario and zero for a hollow conduit.
Assumption boundaryBefore repair, bridge S is zero. Front speeds in mm/day, transition widths in mm, and scaffold support are numerical assumptions. These are prescribed logistic fronts, not measured migration or angiogenesis solutions.
Axon extension along a synthetic path
v0 is the chosen extension scale, ai is a fixed seeded speed trait, and Ti = √[1 + (dy/dx)² + (dz/dx)²] converts path-length growth to axial advance. Fields are sampled at the tip.
Assumption boundaryThe multiplicative coefficients and synthetic path shape are original assumptions. This is not a molecular growth-cone force balance or chemotaxis model.
Chronic axotomy and initiation
Hi is an assumed competence half-life. Ui and each axon’s onset and speed traits are sampled once from the seed, before time integration.
Assumption boundaryA noninitiating modeled axon is not declared dead. Changing the seed samples hypothetical cohort variability, not an empirical distribution of patients.
Local post-arrival maturation
ai(x) is that axon’s first arrival time at x. Lm is the assumed myelination lag and τm is the chosen maturation time constant. This clock is independent of repair-state support loss.
Assumption boundarym is a maturation proxy, not a g-ratio, sheath thickness, or measured conduction velocity. The whole-cohort conduction proxy averages target-arrival maturation multiplied by a seeded caliber factor.
Target-condition diagnostic
Ht is an assumed target-receptivity half-life. This displayed diagnostic is separate from the growth equations and does not gate arrival counts.
Assumption boundaryThe diagnostic has no axon-contact rescue feedback and is not a neuromuscular-junction count, muscle strength, irreversible deadline, or clinical prognosis.
Space, events and time integration
Heun’s explicit predictor–corrector method integrates local reactions and axon advance. The internal maximum step is 0.25 day; the interface also offers 0.1 and 0.05 day refinement. Field nodes concentrate in the gap and near the distal stump, preserving the gap boundary when the target is far away. Spatial averages account for nonuniform cell widths.
Daily output is supplemented by initiation, distal-entry and target-arrival event times. Arbitrary replay-time population counts use individual events, so an interpolated graph does not become a fractional count of neurons. Trajectory storage is distinct from the solver step; its local simplification tolerance is not a guarantee of global geometric accuracy.
Schwann support describes still-denervated pathways. Reinnervating one representative trajectory does not reset the entire local environment; per-tube reinnervation feedback is absent. Local myelination has its own arrival clock and remains possible when repair-state support has declined.
Relationship to published computational models
Podhajsky and Myers modeled interacting tissue populations using diffusion-reaction equations. Laranjeira and colleagues used a random-walk model of neurite growth in engineered constructs. Those studies establish computational precedents; the present local reactions, prescribed fronts and seeded curves do not reproduce either framework or inherit their calibration. Podhajsky–Myers · Laranjeira et al.
4. Parameters and evidence class
Human-scale and rat-scale choices identify context. Changing the species label does not silently rescale coefficients. Human and rodent evidence stays separately identified; neither setting is a validated species fit.
| Parameter | Initial value | Class | Interpretation |
|---|---|---|---|
| Injury gap | 8 mm | geometry | Distance between cut ends; crush and apposed direct repair use zero gap. |
| Distal stump to target | 70 mm | geometry | Distance beyond the distal end of the bridge; total regeneration distance is gap + distal distance. |
| Nerve radius | 1 mm | geometry | Idealized nerve envelope; displayed fiber diameters are enlarged. |
| Repair delay | 0 days | scenario | Measured from injury; recipient support deterioration begins at injury, not at repair. |
| Observation period | 180 days | scenario | Elapsed days since injury, including repair delay. |
| Unimpeded elongation | 1 mm/day | evidence-context | User-selected potential speed. Local environment and path tortuosity reduce axial advance; no universal human speed is assumed. |
| Post-repair initiation lag | 4 days | assumption | Median-scale onset parameter with fixed per-axon heterogeneity; not a clinical estimate. |
| Track alignment | 0.8 index | assumption | Controls geometric guidance and rotational mismatch; dimensionless exploratory parameter. |
| Repair Schwann support | 0.85 index | assumption | Supportive repair state, distinct from Schwann-cell number and remyelination capacity. |
| Vascular support | 0.85 index | assumption | Relative vascular support; no oxygen transport or blood-flow solver is implied. |
| Fibrosis burden | 0.2 index | assumption | Local inhibitory fibrosis proxy, not a CNS glial scar or a measured collagen concentration. |
| Repair-support half-life | 120 days | assumption | Exploratory decay after a 28-day plateau; not fitted to human or rat outcomes. |
| Axotomy competence half-life | 180 days | assumption | Separate exploratory chronic-axotomy effect on initiation, not modeled neuronal death. |
| Target receptivity half-life | 240 days | assumption | Separate denervation-age proxy; not strength, sensation or inevitable irreversible loss. |
| Local myelination lag | 7 days | assumption | Local clock starts when an axon reaches each position, not at injury. |
| Myelin maturation time constant | 35 days | assumption | Local exponential maturation after axon arrival, not a fitted human recovery law. |
| Representative parent axons | 128 axons | numerical | Representative cohort; never an anatomical nerve axon count. Sprouts are not counted as new neurons. |
| Solver time step | 0.25 days | numerical | Second-order predictor-corrector; internal maximum step 0.25 day. |
| Random seed | 2048 | numerical | All random traits are sampled once; no new random values are drawn during integration. |
Why each mechanism is present
Axon extension speed
Literature scale; adjustable assumptionHuman distance/time estimates are of order millimeters per day; Buchthal and Kühl inferred 1.5–2.0 mm/day from sensory-response return after median-nerve suture. Brushart et al. measured approximately 3.3 mm/day in their rat motor-axon experiment.
A chosen speed scale is modified by the modeled local environment; it is not a fitted patient or species distribution. [15] [17] [13]
Repair-entry latency
Mechanism supported; numerical assumptionAxons do not all begin successful regeneration at the same instant; an intervention can alter the onset without altering elongation speed.
A separate, heterogeneous start delay controls entry into the repair rather than hiding all delay in a slower speed. [13]
Wallerian degeneration and debris clearance
Mechanism supported; phenomenological kineticsHuman distal electrical responses disappear over days; Schwann-cell and macrophage clearance involves multiple cellular pathways.
Distal integrity and debris have distinct time courses. Neither normalized field is an assay value. [1] [3] [4]
Repair Schwann support and persistence
Mechanism supported; uncertain time constantRepair-state loss with prolonged denervation is documented in rodents and human tissue, on different time scales and with different measurements.
The support index and its persistence time are explicit assumptions. Mouse c-Jun protein levels are not directly converted to human growth probabilities. [2] [7] [10] [11]
Vascular support
Mechanism supported; normalized surrogateMacrophage-induced vessels guide Schwann migration across rodent nerve bridges.
A dimensionless parameter modifies the local growth environment. It is not blood flow, oxygen tension, or an angiogenesis calculation. [5]
Guidance, alignment, and scar resistance
Mechanism supported; numerical assumptionsRepair Schwann cells form longitudinal tracks, and the architecture of the bridge and distal pathways influences regeneration.
Synthetic fascicles, guidance vectors, and path perturbations encode selected geometry. They are not reconstructed histology or a continuum mechanics model. [6] [5] [12]
Original fascicle matching
Mechanism supported; simplified routing ruleAxons reaching a distal pathway can still be misdirected; anatomical continuity alone does not guarantee a useful target connection.
The original and destination synthetic fascicle indices are compared. Motor/sensory labels are descriptive; this check does not establish correct organ reinnervation or functional recovery. [12] [9]
Target viability and maturation
Mechanism supported; exploratory surrogateLong denervation changes the target environment, while structural regeneration, remyelination, conduction, and force recover differently.
Target condition and post-arrival maturation are separate assumptions. These are neither an NMJ count nor a clinical prognosis. [9] [7] [14] [17]
Repair geometry and material choice
Scenario design, not comparative-effectiveness evidenceClinical repair results depend on the exact nerve, gap, intervention, timing, and endpoint.
Generic suture, graft, and conduit scenarios alter model geometry and assumptions. They do not rank commercial devices or surgical treatments. [16]
Axon population, seed, and repeated runs
Computational designA finite representative population is used so trajectories can be inspected in a browser.
Repeatability and a spread across seeds describe numerical stochastic variability. They do not supply population confidence intervals or capture uncertainty in biological assumptions.
Published computational precedents
Methodological context; no transferred calibrationPublished regeneration frameworks include diffusion-reaction populations and random-walk neurite trajectories coupled to engineered-tissue geometry.
This release instead integrates local reactions along prescribed spatial fronts and advances axons on seeded synthetic curves. Citing earlier frameworks does not make it a reproduction or an independently validated extension of them. [18] [19]
5. Numerical verification and reproducibility
Verification asks whether the implemented equations behave correctly. Biological validation additionally requires fitting and testing against independent, endpoint-specific experiments. That validation has not been performed.
Recorded verification: 36 named numerical tests passed for engine cd5fbd17d1dd. Recorded environment: v24.19.0 on linux. These measurements describe numerical consistency, not agreement with biological outcomes.
| Check | Recorded result | Meaning |
|---|---|---|
| Ideal constant-speed position | 2.70e-13 mm maximum error | Closed-form solution for the test-only uniform ideal limit. |
| 0.25 versus 0.05 day integration | 0.0030 mm maximum endpoint difference | Same seeded default experiment, finer integration. |
| Shared target-arrival events | 0.0101 day maximum difference | Parents arriving in both integration runs. |
| Large numerical cohort | 2.96 s for 512 parents over 730 days | One Node.js run at 0.1 day steps and a 500 mm distal path. Browser runtimes vary. |
Recorded timestep-refinement endpoints
| Step / days | Distal entry | Target arrival | Fascicle match | Maximum position difference / mm |
|---|---|---|---|---|
| 0.25 | 128 | 96 | 77 | 0.00304 |
| 0.1 | 128 | 96 | 77 | 0.00039 |
| 0.05 | 128 | 96 | 77 | 0.00000 |
Exact limiting case
Uniform ideal-reference tests compare x = v(t − lag), entry and target-arrival times with closed-form values.
Conservation & bounds
Disconnected intact material transfers into debris and cleared material. State bounds, nonnegative advance and parent bookkeeping are checked.
Event correctness
No target arrival precedes entry, and no growth precedes initiation. Zero speed and unrepaired injuries cannot receive false arrival credit.
Refinement
Timestep and spatial-grid refinements test consistency. Small gaps remain resolved when the target is hundreds of millimeters away.
Independent clocks
Repair, denervation history, initiation, local arrival, myelin maturation and target receptivity remain distinct.
Seeded replay
Random traits are sampled before integration. Timestep changes do not redraw the cohort.
Exports and comparisons
JSON contains the model version, normalized configuration, seed, full trajectories, daily measurements, fields, assumptions and diagnostics. CSV contains the daily aggregate series. Imports require an exact version match, preserve the supported configuration and recompute locally. Saved result objects are ignored; imported comparisons, ensembles and display state are not restored.
Repeated runs vary the random seed at fixed assumptions. The 10th–90th percentile band describes a finite stochastic ensemble. It is not patient variability, measurement uncertainty, coefficient uncertainty or a biological confidence interval. Sensitivity sweeps hold the seed fixed and vary one parameter across five values; the resulting differences are conditional on the chosen equations and the fixed observation horizon.
6. Anatomical display and unresolved biology
The nerve is procedural: an idealized envelope with three fascicles, representative axons, Schwann spindles, vascular guides and myelin sleeves. This is neither electron-microscopy reconstruction nor patient anatomy. The model’s millimeter coordinates determine tip positions. Transverse scale and fiber diameter are enlarged with explicit labels and can be returned to unity.
The 3D view uses an injury window; the full-path plot includes the entire distal distance. Larger numerical cohorts keep every parent in analysis while the renderer uses a representative subset. Selecting any parent makes it inspectable. A labeled Canvas2D fallback uses the same paths, fields, event clocks and local myelin on devices without WebGL.
- Mechanism-informed exploratory model; not calibrated or clinically validated.
- Synthetic nerve geometry and representative axons; not patient-specific anatomy or a complete nerve census.
- Normalized environmental states; no molecular concentration, oxygen transport, immune lineage, or gene-regulatory solver.
- Structural arrival, target matching, and maturation are distinct; none is a clinical recovery probability.
- Human-scale and rodent-scale scenarios are separate parameter hypotheses, not an interspecies conversion.
- No collagen mechanics, surgical tension calculation, neuroma pain model, motor-unit recruitment, muscle force, or cortical adaptation.
- Conduction-related readouts are surrogates; no biophysical action-potential or extracellular recording calculation.
- Changing random seeds samples model stochasticity; it does not establish empirical confidence limits.
- Replay and local-myelin positions interpolate adaptively simplified trajectories; exact initiation and crossing events are retained.
Predictive validation would require selecting a species, nerve, injury, repair and endpoint; obtaining raw longitudinal observations; estimating identifiable coefficients; and testing held-out data with parameter uncertainty. Additional visual detail alone does not establish that validation.
7. Primary-source evidence dossier
Each original study is identified by biological context, its contribution to this model, and the boundary on its use. Interventions described in these papers are not automatically implemented in the simulator.
1992
Wallerian degeneration in human nerves: serial electrophysiological studies
Reported findingDistal motor responses fell by half at 3–5 days and disappeared by day 9; sensory responses fell by half around day 7 and disappeared by day 11. Distal stump length affected timing.
Use in this modelSeparates injury from the subsequent loss of distal axon integrity and illustrates why degeneration has a finite time course.
BoundaryA small electrophysiological series, not a direct measurement of myelin-debris clearance or a universal degeneration clock.
2012
c-Jun reprograms Schwann cells of injured nerves to generate a repair cell essential for regeneration
Reported findingSchwann-cell c-Jun controls an injury-induced repair program; removing it disrupts trophic support, regeneration tracks, and successful nerve repair.
Use in this modelMotivates a repair-support state distinct from mature myelination and a dependence of growth on the local Schwann environment.
BoundaryThe simulated support index is not a c-Jun concentration, gene-expression measurement, or gene-regulatory model.
2015
Schwann cell autophagy, myelinophagy, initiates myelin clearance from injured nerves
Reported findingInjury activates Schwann-cell autophagy of myelin, regulated in part by JNK/c-Jun; disrupting autophagy delays myelin breakdown.
Use in this modelSupports gradual debris clearance and the role of repair Schwann cells in preparing the distal pathway.
BoundaryThe model compresses several clearance pathways into a normalized state and does not reproduce autophagosome or lipid kinetics.
2017
Schwann cells use TAM receptor-mediated phagocytosis in addition to autophagy to clear myelin in a mouse model of nerve injury
Reported findingSchwann cells and recruited macrophages cooperate in myelin removal; Schwann cells use TAM-receptor phagocytosis as well as autophagy.
Use in this modelGrounds the interaction between debris, macrophage activity, and Schwann support rather than assigning all clearance to one cell type.
BoundaryMacrophage activity is an aggregate index, not a cell count or an M1/M2 classification.
2015
Macrophage-induced blood vessels guide Schwann cell-mediated regeneration of peripheral nerves
Reported findingHypoxia-sensing macrophages induce polarized vessels through VEGF-A. Migrating Schwann cells use those vessels as tracks, and disrupting vessel orientation disrupts repair.
Use in this modelMotivates separate vascular support and aligned guidance in the repair bridge.
BoundaryPerfusion is a normalized support parameter. The simulator does not solve oxygen transport, VEGF chemistry, blood flow, or endothelial sprouting.
2017
After nerve injury, lineage tracing shows that myelin and Remak Schwann cells elongate extensively and branch to form repair Schwann cells, which shorten radically on remyelination
Reported findingAdult myelin and Remak Schwann cells form elongated, branched repair cells after injury and shorten again when they remyelinate regenerated axons.
Use in this modelGrounds longitudinal Schwann tracks and the visual distinction between repair cells and new myelin segments.
BoundaryRendered cells and tracks are synthetic. They are not segmented from the electron-microscopy dataset.
2000
Effects of short- and long-term Schwann cell denervation on peripheral nerve regeneration, myelination, and size
Reported findingDenervation of four weeks or less did not reduce regeneration in this paradigm; longer denervation reduced axon recruitment. Atrophic Schwann cells still remyelinated axons that reached them.
Use in this modelSeparates the ability to support axon entry from the ability to remyelinate an axon after arrival.
BoundaryNeither four weeks nor the reported decline is a human deadline. Local reinnervation can restore a myelinating phenotype.
1995
Contributing factors to poor functional recovery after delayed nerve repair: prolonged axotomy
Reported findingProlonging the interval without a target connection reduced the number of motoneurons that successfully regenerated, even when the distal pathway was freshly denervated.
Use in this modelMotivates a neuronal regenerative-competence factor separate from distal Schwann support and target condition.
BoundaryA failure to regenerate is not equivalent to neuronal death. The model does not infer surviving neuron counts from stalled trajectories.
2011
The basis for diminished functional recovery after delayed peripheral nerve repair
Reported findingThe neuron, distal Schwann pathway, and muscle make distinguishable contributions to delayed repair. Surviving axons can enlarge their motor units, so axon number and muscle force need not change proportionally.
Use in this modelKeeps arrival, correct target routing, maturation, and target viability separate rather than reporting one recovery percentage.
BoundaryThe simulator has no motor-unit recruitment, compensatory sprouting, muscle mechanics, or measured force prediction.
2021
Failures of nerve regeneration caused by aging or chronic denervation are rescued by restoring Schwann cell c-Jun
Reported findingDistal-stump c-Jun protein declined to about 40% of its one-week level after ten weeks of denervation. Maintaining Schwann-cell c-Jun rescued impaired regeneration in the tested models.
Use in this modelSupports time-dependent loss of repair support and sensitivity experiments on the assumed support-persistence time.
BoundaryA measured protein fraction is not the same as regeneration probability. The ten-week mouse result is not used as a calibrated human half-life.
2020
Characterising cellular and molecular features of human peripheral nerve degeneration
Reported findingHuman injured nerves increased repair-associated c-Jun and p75NTR and reduced EGR2. Repair markers declined with long denervation; c-Jun approached or fell below uninjured levels around 500 days in the sampled data.
Use in this modelProvides human evidence that repair-support deterioration is relevant beyond rodents.
BoundaryHeterogeneous cross-sectional tissue samples do not establish a universal time constant, a sharp recovery deadline, or a patient-specific prognosis.
1993
Motor axons preferentially reinnervate motor pathways
Reported findingMotor axon collaterals can enter both motor and sensory pathways. Selectivity emerges partly through collateral pruning and pathway properties retained after degeneration.
Use in this modelMotivates separating anatomical arrival from route matching. The implemented check is whether an axon returns to its original synthetic fascicle family.
BoundaryMotor/sensory labels do not implement molecular target recognition. A fascicle match is not proof of correct organ reinnervation; collateral competition, pruning, and central relearning are absent.
2002
Electrical stimulation promotes motoneuron regeneration without increasing its speed or conditioning the neuron
Reported findingMeasured elongation rates were 3.38 ± 0.58 mm/day with stimulation and 3.27 ± 0.90 mm/day without it. Stimulation improved the onset of regeneration rather than elongation speed.
Use in this modelDemonstrates why repair-entry lag and subsequent extension speed are separate model terms; supports a rodent-scale speed reference.
BoundaryThese rates and treatment effects apply to the experimental protocol. They are not a universal rat rate or a human treatment prescription.
2012
The relationship between nerve conduction velocity and fiber morphology during peripheral nerve regeneration
Reported findingMotor conduction velocity and total fiber diameter increased during recovery. Axon diameter or g-ratio alone did not track recovery consistently, and internodes remained short.
Use in this modelMotivates a delayed maturation state and cautious conduction-related readouts after structural arrival.
BoundaryA maturation index or diameter-based speed estimate is not a simulated compound action potential, muscle response, or validated functional score.
1947
Rate of regeneration in human peripheral nerves: analysis of the interval between injury and onset of recovery
Reported findingClinical observations related injury-to-recovery intervals to nerve regeneration distance, providing a historical basis for distance-and-time estimates.
Use in this modelSupports displaying extension speed in mm/day and separating the travel-distance calculation from time to useful recovery. The human-scale default is an illustrative scenario.
BoundaryThe familiar approximately 1 mm/day rule is an order-of-magnitude heuristic, not an exact constant or a calibration of this simulator.
2023
A multicenter prospective randomized comparison of conduits versus decellularized nerve allograft for digital nerve repairs
Reported findingThe trial compared specific collagen conduits and processed nerve allografts in 5–14 mm and 15–25 mm digital-nerve gaps.
Use in this modelEstablishes that repair evidence is specific to nerve type, gap length, device, and outcome; motivates separate generic repair scenarios.
BoundaryGeneric simulated conduits and grafts are not the tested products. No trial success rate, motor-nerve efficacy, or universal critical-gap threshold is inferred.
1979
Nerve conduction, tactile sensibility, and the electromyogram after suture or compression of peripheral nerve: a longitudinal study in man
Reported findingAfter median-nerve suture, sensory potentials returned at three to four months, corresponding to inferred advance of 1.5–2.0 mm/day. Electrical response properties and tactile sensibility recovered on different time courses.
Use in this modelAdds a measured human distance/time reference and supports distinguishing arrival, electrical maturation, and sensory recovery.
BoundaryThe suture observations included three complete sections and one partial section. Response-return timing is not direct growth-cone tracking or a universal human speed distribution; no coefficients or recovery forecasts are fitted to these cases.
1995
A diffusion-reaction model of nerve regeneration
Reported findingA five-population diffusion-reaction framework represented interacting regenerative processes, with simulated vascular advance preceding neurite growth.
Use in this modelProvides historical computational context for separating supportive tissue processes from neurite advance.
BoundaryThis laboratory uses local reaction equations and prescribed spatial fronts without spatial diffusion. It does not reproduce this published model, inherit its coefficients, or acquire validation from its reported agreement with experiments.
2022
In silico framework to inform the design of repair constructs for peripheral nerve injury repair
Reported findingA random-walk neurite-growth framework explored collagen hydrogels containing biomaterial fibers, optimized fiber geometry, and assessed robustness through sensitivity analysis.
Use in this modelProvides a computational precedent for inspecting how explicit geometric assumptions affect growth and for reporting sensitivity analyses.
BoundaryThe present simulator uses predetermined synthetic curves and its own support laws. It does not implement the published random-walk or durotaxis model, use its fitted parameters, or claim its validation.