Invisible Customisation - A Data Driven Approach to Predictive Additive Manufacture Enabling Functional Implant Personalisation

点击翻译按钮获取中文摘要

Additive manufacturing (AM), otherwise known as 3D printing, is enabling the production of medical implants that are customised, in terms of size and shape, to a person's skeleton. Compared with devices of a standard size, these personalised designs fit the patient better and as such offer improved aesthetics and reduce surgery times. While customisation has many benefits, the challenge is to ensure each bespoke device is made to the same quality. This is difficult because the implant shape is completely unique and may be very complex. Currently in an effort to ensure quality, researchers make lots of plain cube test samples using various manufacturing settings and then compare properties before deciding what combination to use for the real implant. This trial and error approach takes a lot of time and may not even produce very predictable devices because the optimisation is not performed on shapes that are representative of real implants. In this project we will make various design features common to medical implants (e.g. curved surfaces, screw holes) and collect key performance data during and post manufacture. By using cutting edge mathematics, we will create a network that allows us to accurately predict which manufacturing settings will produce the best quality for any design shape. This tool will help businesses to standardise production of customised medical devices in a quick and accurate manner that is not dependent on the user's knowledge. Thereby we will open up the advantages of AM to more companies and help existing adopters to meet the standardisation requirements of the impending new Medical Device Regulations. Overall this project aims to better understand the relationships between additive manufacturing settings and implant properties, which will help us to improve the quality of these anatomically personalised devices. Beyond this we plan to create a tool to enable the creation of implants that are not only customised to the size and shape of the patient's skeleton but also two critical functionalities: mechanical strength and cell adhesion. It is known that if an implant is too strong compared with the surrounding native bone this can cause it to fail. As such, developing a way to select manufacturing or design parameters that enable mechanical matching to the patient's skeleton will help implants to last longer and reduce the number of failures. Besides mechanical mismatch, the other biggest threat to bone implants is infection. Our preliminary work has shown that surface roughness directly impacts the ability of cells, mammalian and bacterial, to stick onto AM devices. In this project we will exploit this knowledge to enable users to select manufacturing settings that result in a defined surface roughness that either enables or prevents cell attachment. This novel capability could be used, for example to create implants with a surface that stops bacterial cells from sticking and thus minimises infection risks. There is also potential that this tool could help to improve bonding between the implant and native tissue by recommending manufacturing settings that result in surface topographies that encourage growth of bone forming osteoblast cells. In summary, this project is focused on standardising the way we use 3D printing to ensure the properties of bespoke implants are predictable. This will be achieved by using mathematics to move the AM field away from trial and error. By understanding the relationships between manufacturing settings and key properties, we will create two tools that will enable us to make functionally personalised devices. The ability to predictively and selectively tailor mechanical properties and surface roughness will drive a new generation of implants that last longer and fail less often. Thereby, this project will ultimately improve the lives of millions of people who receive bone implants and help to reduce the associated healthcare costs.