Logo
Munich Personal RePEc Archive

Dynamic Spatial Treatment Effects in Neurotransmitter Diffusion: Applications to Movement Disorders

Kikuchi, Tatsuru (2025): Dynamic Spatial Treatment Effects in Neurotransmitter Diffusion: Applications to Movement Disorders.

[thumbnail of MPRA_paper_126720.pdf]
Preview
PDF
MPRA_paper_126720.pdf

Download (837kB) | Preview

Abstract

Traditional spatial treatment effect methods impose arbitrary boundaries between treated and control regions, obscuring how treatments spread through neural tissue. We develop a continuous functional framework deriving explicit treatment boundaries from diffusion physics, eliminating discretization artifacts while providing testable predictions. Our approach applies partial differential equations to neurotransmitter diffusion, unifying spatial scales from synaptic spillover (micrometers) to volume transmission (centimeters).

We validate using synthetic data calibrated to established neuroscience parameters across five conditions: healthy controls, dystonia, Parkinson's disease, Alzheimer's disease, and acute ischemia. Results demonstrate systematic disease-induced boundary alterations. Dystonia reduces treatment reach by 16.3\% (p $<$ 0.0001), requiring 24\% dose increases for equivalent coverage. Parkinson's shows 31.8\% reduction, ischemia 34.3\%. Spatial decay parameters evolve from 483 mm$^{-1}$(1 hour) to 54 mm$^{-1}$ (72 hours), matching theoretical $1/\sqrt{t}$ predictions.

Compared to difference-in-differences methods, our framework achieves superior fit ($R^2$ = 0.92 vs. 0.76) with explicit boundary detection. Non-parametric approaches achieve higher in-sample fit ($R^2$ = 0.99) but lack physical interpretability and cannot identify boundaries.

Clinical applications include: (1) data-driven determination of injection sites and stimulation parameters, (2) disease-specific dose adjustments based on tissue properties, and (3) treatment time course prediction from early measurements. These advances directly address limitations in deep brain stimulation, botulinum toxin therapy, and drug delivery planning.

Methodologically, we extend spatial causal inference from pollution dispersion and financial networks to neuroscience, demonstrating treatment boundaries emerge naturally from diffusion physics across diverse domains.

Atom RSS 1.0 RSS 2.0

Contact us: mpra@ub.uni-muenchen.de

This repository has been built using EPrints software.

MPRA is a RePEc service hosted by Logo of the University Library LMU Munich.