Researchers at the National Institutes of Health Clinical Center have developed MRISegmenter, an automated segmentation tool that accurately and robustly segments over 60 abdominal organs and structures on T1-weighted MRI, addressing the longstanding clinical and research need for reliable, large-scale organ segmentation in abdominal imaging. The development of MRISegmenter involved the creation of a comprehensive T1-weighted abdominal MRI dataset and the training of a three-dimensional deep learning model, which demonstrated impressive performance on internal and external datasets, indicating generalizability and robustness across different patient populations and imaging protocols. The release of MRISegmenter, alongside its comprehensive training dataset, at MRISegmenter GitHub Repository is an important milestone for the radiology community, as it has the potential to facilitate more efficient workflows, improve reproducibility in research studies, and lay the groundwork for advanced radiomics and machine learning applications.

Source: Radiology