How can we replicate the unique mechanical functions of natural materials like wood, turtle shells, and spider silk for orthopaedic restoration applications? A novel answer lies in the recent work led by the University of Illinois Urbana-Champaign.

Material science, utilising insights from engineering, chemistry, and physics, studies the peculiarities and behaviours of all types of natural materials, aiming to optimise the structure of existing synthetic ones and design new ones. The myriad of natural materials includes wood, spider silk, turtle shells, bird feathers, seashells, and animal bones, to name a few examples.

We are practically surrounded by natural materials. «They are the main constituents of plant and animal bodies and serve a variety of functions, the most important of which is mechanical». Materials like nacre and spider silk possess characteristics that make them suitable for engineering applications requiring mechanical resilience [source: “Three-Dimensional-Printing of Bio-Inspired Composites” – Journal of Biomechanical Engineering].
Regarding mechanical function, in the case of wood and bones, this results directly from their being «fibrous composites with hierarchical structure, » able to adapt functionally «at all hierarchical levels» [source: “Nature’s hierarchical materials” – Progress in Materials Science].

What unites the variety of natural materials are their irregular architectures on the microscale, characterised by «disorder and non-uniformity» which confer them interesting functionalities, with mechanical stress modulation being crucial for elasticity and maintaining balance. «Mechanical stress» refers to «the intensity of force exerted on a small area» [source: “Mechanics of Materials: Stress” – Boston University].

A research team led by the University of Illinois has devised a generative computational paradigm, through which they defined a virtual process for designing materials that mimic the structure and characteristics of those found in nature, even programming their mechanical stress modulation.
Once designed and 3D printed, the bio-inspired material was tested in a biomechanical evaluation of a synthetic fractured human femur model. The effectiveness of the new material’s mechanical functions suggested its potential future role in supporting damaged bones, thereby stimulating their regeneration.
In a potential future scenario where the bio-inspired material with optimised mechanical stress modulation is adopted by orthopaedic surgeons, patients could benefit from quicker recovery without prolonged immobilisation and without the risk of chronic pain and micro fractures around the metal prosthesis.

Mechanical properties of natural materials

The use of natural materials, selected for their mechanical performance, is widespread. Turtle shells, for example, owe their robust mechanical properties to their well-organised and stratified structure. This structure is composed of «a mixture of minerals such as calcium phosphate and calcium sulphate, dispersed within keratin, forming a series of nanocomposite plates stacked in an orderly and compact manner. » The impressive tensile and compressive performance of turtle shells has long been studied as an inspiration for developing biomimetic solutions for advanced multifunctional materials, particularly artificial armour [source: “The hierarchical structure and mechanical performance of a natural nanocomposite material: The turtle shell” – Colloids and Surfaces A: Physicochemical and Engineering Aspects].

Spider silk threads also boast significant mechanical qualities, including toughness and elasticity – attributed to specific protein structures – which surpass those of other fibrous materials, «making them superior to Kevlar and steel». These properties make spider silk suitable for innovative materials in robotics, the medical field (artificial muscles), and cosmetics [source: “Recombinant Spider Silk: Promises and Bottlenecks” – Frontiers in Bioengineering and Biotechnology].

Additionally, spider silk is being explored as a potential future alternative to plastic, with a bio-based material created by bonding wood cellulose fibres and silk proteins found in spider webs [source: “Biomimetic composites with enhanced toughening using silk-inspired triblock proteins and aligned nanocellulose reinforcements” – Science Advances].

Natural materials for prosthetic orthopaedics: the example of wood

Wood is increasingly referenced as a sustainable and biocompatible material for prosthetic orthopaedics and bone repair, as an alternative to the metallic implants commonly used for internal fracture fixation, which often require multiple surgeries and are at risk of rejection. «So far, only a few species of wood have been examined, but various preparation techniques, such as boiling in water or preheating ash, birch, and juniper woods, have been proposed», explains a team from Riga Stradins University in Latvia, authors of a 2023 study on wood as a potential renewable material for bone implants [source: “Wood as Possible Renewable Material for Bone Implants – Literature Review” – Journal of Functional Biomaterials].

Research has since shifted to charred wood and cellulose-derived wood scaffolds, from which charred wood and cellulose scaffolds are derived and combined with other materials (silicon carbide, hydroxyapatite, and bioactive glass) to improve biocompatibility and mechanical durability. The team observes that «in all studies conducted so far, wood implants have provided good biocompatibility and osteoconductivity, due to the material’s porous structure».

Another notable example for prosthetic orthopaedics is an interesting Italian project by the Institute of Science, Technology and Sustainability for Ceramics (CNR-ISSMC) in Faenza, described in “Ceramics with the signature of wood: a mechanical insight” (Materials Today Bio, January 2020 issue). This project devised a novel physico-chemical process to manufacture ceramics from rattan wood. The resulting material features a nanostructure that retains elements inherited from rattan, including its woody essence and porosity, yielding mechanical characteristics that generate resistance and elastic rigidity for applications such as bone replacement «where the ceramic piece is subjected to a load with a preferred direction».

Modulation of bio-inspired, programmed and optimised mechanical stress

A team of researchers from the University of Illinois Urbana-Champaign, USA, in a recent study presented in “Modulate stress distribution with bio-inspired irregular architected materials towards optimal tissue support”, published in Nature Communications on 21 May 2024, offers a critical reflection on the mechanical performance of natural materials. They observe that although these performances are qualitatively high, the relationship between mechanical stress modulation and the irregular, heterogeneous structures characteristic of natural materials remains unclear, primarily due to the complexity of investigating it. This means it is not easy to determine which structures possess a greater or more versatile capacity for mechanical stress modulation.

From this arises the authors’ objective to focus on the relationship between material structures and their physical properties. They aim to achieve this using a generative computational framework, designed to guide a virtual design process for materials with heterogeneous and disordered microstructures – similar to those of natural materials – by controlling, programming, and optimising their mechanical stress modulation, a key mechanical function of natural materials particularly suitable for advanced applications.

The Generative Computational Framework

The framework developed by the team involves four steps. The first involves creating a database containing the basic characteristics of natural materials, achieved through sampling 200 combinations of four fundamental elements (strength, stiffness, elasticity, tension), represented virtually in both 2D and 3D.

The second step involves developing a machine learning model – trained with the PyTorch package – tasked with analysing the database content and, based on this, predicting the mechanical stress modulation of individual microstructures.

The optimisation of macroscopic topology is the phase of the generative computational method aimed at establishing a formulation and parameterisation for mechanical stress modulation, which must be considered in the design process.

Finally, a design process simulator is employed, whose function is to virtually generate materials with irregular architecture, heterogeneous microstructures, mimicking existing natural materials, with precise mechanical stress modulation. The simulator used in this study, the research group notes, features a heterogeneous distribution of constituent elements and, compared to the original program, «is based on input frequency combinations founded on intuition. These aspects allow for optimal microstructural layouts and customised material propertiescrucial for precise mechanical stress modulation».

Schema che illustra il workflow del framework computazionale generativo: a) creazione di un database contenente una serie di combinazioni delle caratteristiche base dei materiali naturali, b) sviluppo di un modello ML per la previsione della modulazione dello stess meccanico dei materiali, c) ottimizzazione della topologia macroscopica, per determinare le variabili di progettazione ottimali d) simulazione del processo di generazione di materiali dall'architettura irregolare, ispirati a quelli naturali [fonte: “Modulate stress distribution with bio-inspired irregular architected materials towards optimal tissue support” - Nature Communications -].
Workflow of the generative computational framework: a) creation of a database containing a range of combinations of natural material characteristics, b) development of an ML model to predict the modulation of mechanical stress in materials, c) optimisation of the macroscopic topology to determine optimal design variables, d) simulation of the material generation process with irregular architecture inspired by natural materials [source: “Modulate stress distribution with bio-inspired irregular architected materials towards optimal tissue support” – Nature Communications –].

The production of the bio-inspired material – designed with a controlled distribution of mechanical stress modulation – was achieved through 3D printing using a layer-by-layer polymerisation process.

Application in femoral prosthetic orthopaedics: bio-inspired mechanical stress modulation

The 3D-printed resin prototype of the bio-inspired material was subsequently tested by the US university researchers in a femoral prosthetic orthopaedics application, specifically for repairing a synthetic model of a fractured human femur.

«During the biomechanical evaluation tests of the fractured limb, this type of optimised material demonstrated versatile mechanical stress modulation due to various forces exerted on a specific area of the femur», the authors explain. This modulation suggests that in future real (not synthetic) models of fractured femursbone regeneration could be stimulated through the so-called “shear stress,” which is the natural internal tension the bone experiences even with minimal effort, such as during leg movement.

Conversely, traditional metal plates and intramedullary nails, commonly used in femoral prosthetic orthopaedics for fractures, hinder this natural internal tension, compromising bone elasticity. If the bone substance of a healthy femur is characterised by a uniform modulus of elasticity throughout the length of the bone, fractures cause a concentration of forces at the fracture’s apex, increasing the risk of further propagation.

Prosthetic orthopaedics with bone tissue regeneration

«Traditional methods for repairing a fractured femur involve fixing a rigid plate around the fracture with screws. However, due to the disparity in rigidity between the femur and the plate, there is a negative shielding effect against any internal force, significantly reducing the bone tissue’s elasticity. This leads to further patient immobilisation, with the risk of chronic pain and the development of a fracture around the implant area, known as a ‘periprosthetic fracture», the authors highlight.

A potential future solution to the problem of bone elasticity loss following a fracture and the subsequent fixation of metal prostheses within the bone could come from the application of bio-inspired materials with optimised stress modulation directly at the fracture site. Initial biomechanical tests on a synthetic human femur model suggest this approach.

This type of support would transmit a compressive force to the fractured bone, closely mimicking the natural force, while imparting minimal internal tension (shear stress) between the bone (in this case, the femur) and the support itself. «As a result, this shear stress, inducing perpendicular micro-movements to the pre-existing fracture, would restore sufficient elasticity and stimulate bone regeneration», the researchers emphasise.

Glimpses of Futures

The described work has led to the development of an innovative methodology that utilises computational technologies, including AI techniques such as machine learning, to control, programme, and optimise the most intriguing mechanical function of natural materials. This is aimed at achieving elasticity and balance in synthetic materials for advanced applications, such as prosthetic orthopaedics and bone regeneration, regardless of the fracture site.

Let’s now attempt to foresee possible future scenarios, using the STEPS matrix to predict the impacts that the evolution of this methodology could have from social, technological, economic, political, and sustainability perspectives.

S – SOCIAL: in the future, with the advancement of the generative computational framework developed by the American team, alongside successful laboratory test results on fractured bone models in animals, the adoption of bio-inspired material with optimised mechanical stress modulation would bring a dual benefit to prosthetic orthopaedics. Doctors would avoid performing multiple surgeries in the area where plates, nails, and screws have been fixed, reducing the risk of infections or rejections, thereby saving resources, effort, and time. Patients, on the other hand, would experience faster recoveries without prolonged immobilisation and without the risk of chronic pain and periprosthetic fractures. This is because, instead of metal, their bodies would contain a biocompatible material that, crucially, can restore a certain degree of elasticity to the bone affected by trauma, promoting its regeneration.

T – TECHNOLOGICAL: the presented method is just the beginning of a broader effort by the University of Illinois Urbana-Champaign, aimed at different types of prosthetic orthopaedics applications in the future, wherever enhanced and precise mechanical stress modulation is needed. The authors also highlight that the proposed technique is effective for the control, programming, and optimisation of mechanical stress modulation in any type of material. «The key – they comment – is the architecture of the material in question, its microstructure, and the corresponding mechanical properties. These are the factors that make the applications of the generative computational framework virtually limitless». In the future, its evolution will lead to the development of a neural network with a deeper architecture for predictive analysis not only of mechanical stress modulation but also of other characteristics inherent to natural materials.

E – ECONOMIC: bone fractures, particularly femoral fractures, are quite common among the elderly population, mainly due to the deterioration of bone tissue caused by osteoporosis. According to statistics from the International Osteoporosis Foundation (IOF), globally, up to 37 million fractures are recorded annually among people over the age of 55 due to bone fragility, significantly impacting healthcare expenditure due to hospitalisation costs. In our country, based on the latest available data on the direct costs of fractures from 2019, these costs are approximately 5.4 billion euros. In this context, the future use of a bio-inspired material, like the one described, to support prosthetic orthopaedics – capable of reducing the risk of fractures around the metal prosthesis, the risk of rejection of plates, screws, and nails, infections, and multiple surgeries – would result in shorter hospital stays and thus reduce the associated costs for the National Health Service.

P – POLITICAL: in a future scenario, the use of an unconventional prosthetic support in orthopaedics, different from metal prostheses and inspired by the microstructures of natural materials (currently, the prototype is made of biocompatible resin), such as the one designed by the US research group using a generative computational method, would require legislative attention regarding safety aspects, as it is currently only tested in the laboratory. We should recall that from 26 May 2021, the new EU Regulation 2017/745 on medical devices, including breast and hip/femoral prostheses, has been in force, placing particular emphasis on the safety and performance requirements of medical devices marketed within the Union, with specific obligations and responsibilities for manufacturers, the supply chain, and healthcare institutions. This regulation will need to integrate the use of prosthetic implants with bio-inspired mechanical characteristics, reproduced using computational methods, in the future.

S – SUSTAINABILITY: orthopaedics will increasingly adopt prosthetic implants made from more natural and biocompatible materials, which are also elastic and durable, in the name of uncompromising environmental sustainability and a sustainability model that also embraces human health. This respects the body’s natural physiology when it suffers trauma, such as a fracture, and needs an internal prosthesis that does not immobilise it but instead supports bone regeneration, as the new bio-inspired material promises. The only downside, considering the computational framework created by the authors for its design, is the carbon footprint generated by the machine learning system on which it relies, especially during the training phase.

Written by: