In a potential game-changer for affordable magnetic resonance imaging (MRI) machines and future electrified transport systems, scientists have developed the world’s strongest iron-based superconducting magnet using artificial intelligence (AI).

Superconducting magnets are highly valued for their ability to produce strong and stable magnetic fields without significant power consumption. This capability is crucial for technologies such as MRIs, which rely on magnetic fields to generate clear, three-dimensional images of soft tissue.

Researchers from King’s College London and Japan have harnessed machine learning (ML) to create a cost-effective and powerful iron-based superconducting magnet, which could revolutionize the accessibility and affordability of this technology. Dr. Mark Ainslie from King’s Department of Engineering collaborated with experts from Tokyo University of Agriculture and Technology, the Japan Science and Technology Agency, the National Institute for Materials Science, and Kyushu University on this groundbreaking project.

The team succeeded in producing a superconducting magnet with a magnetic field 2.7 times stronger than previously reported. They utilized a new ML system called BOXVIA, which optimized the creation process of superconductors more efficiently than traditional methods. BOXVIA identified performance-enhancing patterns and fine-tuned parameters to design the optimal magnet. This process, which would typically take months, was significantly expedited.

Remarkably, the AI-developed superconducting magnet exhibited larger iron-based crystals within its structure, differing from the uniform crystal structure commonly seen in magnets made without BOXVIA. This variety in crystal size is a testament to the unique capabilities of the AI system.

“Superconducting magnets are the backbone of the future. Not only are they used to image cancers with MRI machines, but they will be vital for electric aircraft and nuclear fusion,” said Dr. Ainslie. He noted that the materials and technology required for traditional copper-based superconductors are expensive, limiting their market penetration.

Using bulk form superconductors, which retain magnetism once magnetized, offers a smaller footprint compared to heavier wire coils. However, fabricating copper-based bulk superconductors can take weeks. The iron-based magnet developed by the team promises a more efficient and cost-effective alternative.

This research lays the foundation for making superconducting magnets powerful enough for industrial applications. “Using artificial intelligence, we’ve produced a cost-effective and scalable alternative using iron, which is a lot easier to work with and opens the door for smaller and lighter weight devices,” Dr. Ainslie explained.

The first iron-based superconductors, made over a decade ago, produced magnetic fields that were not strong or stable enough for widespread use. This breakthrough could make MRI machines cheaper and lead to the development of smaller MRI units suitable for GP offices, thereby enhancing accessibility.

This AI-driven innovation represents a significant leap forward, potentially transforming not only medical imaging but also various applications requiring powerful and efficient magnetic fields.

By Impact Lab