Two researchers from the Institute for Neurosciences, a joint center of the Spanish National Research Council and the Miguel Hernández University of Elche, have developed a revolutionary method to accelerate advanced brain MRI scans. Published in Communications Medicine, the study demonstrates that combining artificial intelligence with computer simulations can reduce scan times by up to 90% while maintaining high diagnostic accuracy. This breakthrough addresses a critical bottleneck in clinical neuroimaging by allowing detailed brain analysis using only a fraction of the data typically required. The core innovation lies in the training methodology of the artificial intelligence models used. Unlike current applications that rely on real patient data, which can be limited and raise privacy concerns, this team trained their neural networks using simulated data generated from the physics of diffusion processes in brain tissue. This physics-based approach enables the model to learn how to reconstruct detailed tissue microstructure from very few resonance images. The researchers emphasize that simulations allow them to generate vast amounts of data on demand without depending on patient availability, thereby eliminating biases associated with traditional clinical datasets. The study specifically targets diffusion-weighted MRI, a technique that non-invasively studies water movement in brain tissue to reveal its microstructure. By training a network on simulated data, the team proved that the system achieves excellent accuracy using just 10% of the standard measurements. In practical terms, this translates to a dramatic reduction in scan duration. A procedure that traditionally takes approximately 40 minutes could be completed in as little as eight minutes. Such efficiency would allow hospitals to treat significantly more patients within the same timeframe, potentially alleviating pressure on facilities with long waiting lists. This advancement holds particular promise for the early diagnosis of neurodegenerative diseases such as Alzheimer's. These conditions often have a preclinical phase lasting up to two decades where no visible symptoms appear, making early detection challenging with existing diagnostic tools developed over 30 years ago. The new method provides more detailed information during this silent phase, improving the ability to detect diseases before they manifest clinically. Furthermore, the technology offers a unique capability to reanalyze MRI data acquired decades ago. Previously limited by the technology of the time, historical scans can now be reinterpreted to extract new, relevant information about neurological conditions. Researchers Silvia De Santis and Maximilian Eggl, who led the work from the Institute for Neurosciences, highlight that reducing acquisition time not only increases efficiency but also expands the types of advanced MRI techniques that can be practically incorporated into clinical workflows. By providing medical staff with greater amounts of clinical information in a shorter period, this simulation-driven AI strategy paves the way for a more accessible and effective future for neuroimaging diagnostics.