How Artificial Intelligence is being used to design the implants of the future

• From design to final verification, the METALIA project is developing innovative tools to address existing challenges in the design, manufacturing, post-processing, operation, and characterization of new personalized shape memory medical devices.

At the forefront of medical innovation, implants have undergone a revolution in recent times thanks to artificial intelligence (AI). Nowadays, these devices are designed and optimized using advanced algorithms.

However, conventional manufacturing techniques do not allow for patient customization.

Therefore, the METALIA team has taken on the challenge of developing new additive manufacturing (3d printing) routes, for implants based on shape memory alloys.

IMDEA Materials Institute is one of the partners of this ambitious project, coordinated by the Advanced Center for Aerospace Technologies (CATEC) and the manufacturer of aerospace components and systems, Egile.

Nitinol: a benchmark alloy for biomedical applications

Nitinol, composed of nickel and titanium, is one of the most widely used shape memory alloys (SMAs) in biomedical applications, especially for the development of self-expanding implants in minimally invasive surgeries.

However, it has always been employed using conventional manufacturing routes. This is where new 3D printing techniques come into play, allowing us to overcome existing limits device manufacture.

Specifically, additive manufacturing will open the door to developing patient personalized implants, resulting in less aggressive treatments with better prognoses.

Within the framework of the METALIA project, IMDEA Materials will study different methods to optimize 3D printing processes, ensuring the correct concentration of each element in the device, and that it maintains its properties.

For these medical devices to serve their function, they must allow cell proliferation around them. However, if the surface roughness of the piece is very high, it affects the material's cytotoxicity.

To address this, several project partners, including Leitat Technological Center, the Polytechnic University of Madrid, Grupo Sevilla Control (GSC), BitMetrics, DLyte, and cubicoff, will focus on developing optimized surface treatments that guarantee good implant integration.

"Tools for the design and calculation of the metallic components to be manufactured, quality control, and failure prediction during additive manufacturing will also be developed", explains Dr. Irene Heras, Principal Investigator at project coordinators CATEC.

In addition to optimizing 3D printing processes, IMDEA Materials will also direct its research towards printing SMAs. Hence, Nitinol has been chosen.

This alloy possesses self-healing and shape memory properties; it can deform at low temperatures, maintain its deformed shape when external force is removed, and recover its original shape when heated above its transformation temperature.

"4D printing means that a material, once printed, can change shape," says Dr. Federico Sket, one of IMDEA Materials' principal researchers involved in the project, along with Dr. Jon Molina. "With this technique, medical devices can be developed to optimally fit the unique shape and size of a person."

Making characterization of 3D-printed materials faster and more affordable

One of the most useful tools for characterizing materials manufactured with complex 3D geometries is X-ray tomography.

This technique provides a digital volume of the object from which, using image processing techniques, relevant information about the manufacturing process or post-processes applied to the material can be extracted.

However, as obtaining a digital volume of sufficient quality requires acquiring a set of radiographs from various angles, this technique is time- and resource-intensive. Generally, the more radiographs and longer exposure times for each radiograph, the better the quality of the digital volume.

Currently, characterizing implants involves taking around 1,500 radiographs. AI techniques can help reduce this burden (either by reducing the radiograph acquisition time or limiting the number of projections or radiographs) while maintaining image quality.

For this purpose, IMDEA Materials researchers are training AI models using thousands of experimental radiographs, both within the METALIA project and in other studies conducted at the institute.

"In the manufacturing of optimal and personalized devices, it is necessary to extract qualitative information, we need to know what is inside the material," explains Dr. Sket. "We also need quantitative information, and this information needs to be extracted automatically."

There are two alternatives to developing these AI tools: either eliminate tomography image 'noise' or, by being able to accurately analyse the piece despite this interference. IMDEA Materials will focus on the first of these objectives, through the development of a database.

METALIA: towards the future of 4D Printing

If an AI-based tool can be developed to accurately characterize properties more effectively in terms of time and resources, it could contribute to the more sustainable implementation of devices manufactured using 4D printing.

Besides the characterization challenges, METALIA aims to protocolise an additive manufacturing technique. It is essential that the alloy manufactured maintains the precise concentration of each element, as small variations can result in significant changes in its characteristics and properties.

"In this four-year project, four research organizations and five companies are participating, fostering an environment that promotes the generation of scientific results towards advanced levels of technological maturity," comments Dr. Heras. "And that is all thanks to a proper transfer between scientific research and technological development."

PLEC2023-010237 (AEI); MIG-20232050 (CDTI)

The project belongs to the "Science and Innovation Missions" program call within the framework of the Transmissions initiative of 2023, an action executed in collaboration between CDTI and the State Research Agency (AEI), coordinating the funding for groups of research and knowledge dissemination organizations and groups of companies collaborating for the joint development of a coordinated R&D action that responds to the challenges identified in the thematic priorities (Missions) defined in the call. The project is funded by the Ministry of Science, Innovation and Universities, the State Research Agency, and the Center for Technological Development and Innovation E.P.E. (CDTI).

Provided by IMDEA Materials