Mechanical Properties of the Meninges: Large Language Model Assisted Systematic Review of over 25,000 Studies

TL;DR

This systematic review, aided by large language models, comprehensively analyzes the mechanical properties of meningeal layers, revealing significant underestimations in current models and highlighting key gaps for future research to improve brain injury simulations.

Contribution

The study introduces a large-scale, LLM-assisted systematic review of over 25,000 studies, providing the most comprehensive data on meningeal mechanics and identifying critical gaps in current knowledge.

Findings

Elastic modulus of pia mater is underestimated by an order of magnitude.

Meningeal layers exhibit nonlinear rate dependence varying with species, age, and location.

Current FE models often oversimplify meningeal properties, affecting injury prediction accuracy.

Abstract

Accurate constitutive models and corresponding mechanical property values for the meninges are important for predicting mechanical damage to brain tissue due to traumatic brain injury. The meninges are often oversimplified in current finite element (FE) head models due to their complex anatomy and spatially-variant mechanical behavior. This study performed a systematic review (SR) on the mechanical properties of each individual layer of the meninges to obtain benchmark data for FE modeling and to identify gaps in the current literature. Relevant studies were filtered through three stages: a broad initial search filter, a large language model classifier, and manual verification by a human reviewer. Out of over 25,000 studies initially considered, this review ultimately included 47 studies on the dura mater, 8 on the arachnoid mater, and 7 on the pia mater, representing the largest and most comprehensive SR on the mechanical properties of the meninges. Each layer was found to exhibit nonlinear rate dependence that varies with species, age, location, and orientation. This study revealed that the elastic modulus of pia mater most often used in simplified linear elastic FE models is likely underestimated by an order of magnitude and fails to consider directional dependence. Future studies investigating the mechanical properties of the meninges should focus on a wider range of loading rates as well as age effects for the arachnoid mater and pia mater, as these features are relatively understudied and expected to affect the fidelity of FE predictions.