New Model Makes MRI More Accurate and Reliable

This non-invasive imaging technique uses a magnetic field and radio waves to capture images of soft tissues without involving radiation. However, MRI does have some limitations. Movements by the participant, such as blinking, breathing, or other involuntary actions, can cause image blurring and lead to ghost artifacts, which repeat the structures. This can be particularly challenging for children who find it difficult to remain still throughout the scan. As MRI is crucial for diagnosing brain conditions and conducting neurological research, maintaining high image quality is essential. Now, a new MRI model offers enhanced accuracy and reliability for brain structure analysis.

To address these issues and improve brain MRI image quality, researchers at the UNC School of Medicine (Chapel Hill, NC, USA) have developed the Brain MRI Enhancement foundation (BME-X), a model designed to correct motion, improve resolution, reduce noise, and enhance image contrast. A standout feature of this model is its ability to "harmonize" MRI images from various scanners. With many different MRI scanners in use across clinical settings, each with its own imaging parameters, achieving consistent results can be difficult. BME-X can integrate this data and standardize it, providing “harmonized” images suitable for both clinical and research purposes.

In a study published in Nature Biomedical Engineering, the BME-X model was tested using over 13,000 images from diverse patient groups and scanner types. The researchers found that BME-X outperformed other leading methods in addressing body motion, reconstructing high-resolution images from lower-resolution data, reducing noise, and managing pathological MRIs. The model’s strength in harmonizing data positions it to streamline clinical trials and research involving multiple institutions, while also contributing to the development of new, standardized neuroimaging protocols and procedures.

“Imaging quality is important for visualizing brain anatomy and pathology and can help inform clinical decisions,” said Li Wang, PhD, associate professor of radiology. “Our model can perform more accurate and reliable analysis of brain structures, which is critical for early detection, diagnosis, and monitoring of neurological conditions.”