Artificial intelligence (AI) may help reduce the need for contrast dye during MRI scans for brain tumor patients, according to a new study. Contrast dye, specifically a substance called gadolinium, is often injected into the bloodstream to enhance the clarity of MRI images, making it easier to detect tumors. However, the long-term health effects of gadolinium exposure are not fully understood, and the substance has also been found in various water sources, including sewage and drinking water, far from MRI centers. This has raised environmental and health concerns. Researchers at University College London (UCL) created an AI tool that predicts which parts of the brain would appear brighter on MRI scans if contrast dye had been used. The tool was trained using 11,089 MRI scans from over 8,500 patients in the UK, US, Netherlands, and Nigeria. When tested on 1,100 scans without contrast dye, the AI correctly predicted whether a tumor would brighten in 83% of cases. It identified 92% of tumors that did brighten and 74% that did not. The AI performed best in detecting meningioma, a common type of brain tumor, but was less accurate when used on scans of children. The AI model shows promise as a tool to help doctors decide whether to use contrast dye during MRI scans. It could help identify cases where contrast is likely to be needed, allowing for a more efficient scanning process and potentially reducing the need for multiple hospital visits. However, the researchers noted that the model is not yet advanced enough to fully replace contrast-enhanced MRI scans, especially for children. Further development and collaboration with radiologists are needed to improve its accuracy and effectiveness. Dr. Karen Noble, director of research and policy at Brain Tumour Research, welcomed the potential of this AI innovation. She noted that over 100,000 people in the UK are living with a brain tumor or its long-term effects, many of whom undergo regular MRI scans. Reducing the need for contrast dye could help minimize possible side effects and improve patient care. The organization looks forward to seeing how this technology can be integrated into future medical practices to support better treatment decisions for brain tumor patients.