Exhibitors & Products
Events & Speakers

The QI-Digital initiative networks various activities and processes relating to the digitalization of the quality infrastructure (QI). Both the practice and regulation of quality assurance are being rethought across all elements of QI. The initiative addresses existing and new quality issues in a digitalized world that requires sustainability and resilience. The resulting digital solutions should ultimately enable innovative quality assurance in and with institutions and companies. In this context, the international transformation process is also being shaped with a view to global connectivity. With ideas from all stakeholders, QI-Digital is thus making a significant contribution to renewing the competitiveness of the economy - for a Made in Germany and Made in Europe of the 21st century.

When it comes to the currently omnipresent topic of artificial intelligence, the "question of trust" is being asked more and more frequently, especially when it comes to medical matters. For this reason alone, the AI algorithms used there must be thoroughly tested. For the "AI in medicine" pilot project, QI-Digital has therefore developed a special testing platform for the AI-supported interpretation of heart data collected using a magnetic resonance imaging (MRI) scanner. The initiative is now showcasing a mobile prototype of this solution at the stand of the Federal Ministry for Economic Affairs and Climate Protection at HANNOVER MESSE 2024. The project is based on the TraCIM system, which the Physikalisch-Technische Bundesanstalt (PTB) already uses to offer digital metrological services and which is now being further developed for testing AI algorithms. A test process can be run through as an example on the demonstrator at the trade fair.

A new type of low-field magnetic resonance tomograph (MRT) will also be on display in Hanover. Due to its compact size, it could in future come to the patient and not the patient to the MRI scanner. This innovation is possible thanks to the combination of AI algorithms with cost-effective hardware tailored to clinical applications. This is because new deep learning algorithms help to improve the quality of images, which contain slightly more noise in low-field MRIs than in high-field scanners. In deep learning, neural networks are fed with simulated data, optimized with end-to-end training and the results are then compared with real measurements. The networks are also fed with prior physical knowledge in order to tailor them more robustly and efficiently to the respective application. The system will be published under open licenses and fully characterized in order to provide easy access to this complex technology as a reference system for science, industry and ultimately patients.