Abstract
Quantitative imaging is increasingly used for both clinical and research applications in musculoskeletal (MSK) disorders. Its widespread use, coupled with an assortment of modalities and vendors, has led to a diverse range of analysis methods and challenges in reproducing results both within and across centers. Clearly, consensus is needed to establish consistent data organization principles and thus permit the use of standardized image processing pipelines. Here, we-members of the Open and Reproducible Musculoskeletal Imaging Research (ORMIR) community-introduce the ORMIR-MIDS data format, which is derived from the existing Brain and Medical Imaging Data Structures (BIDS and MIDS) and extends them to MSK applications. ORMIR-MIDS comprises both a standard specification and a set of software tools for data conversion and organization. The latter permits the conversion of image data to a standardized format, in an organized folder structure, and produces up to 3 metadata files: important data-processing information, patient data, and complete image header tags to allow conversion back to the original format. The tool currently supports a range of imaging modalities and sub-modalities relevant to the MSK system, with more to come in the future. A suite of test data is also provided to demonstrate the functionality of the software, and the file system structure and associated image files and metadata after conversion of these test data are demonstrated here. With ORMIR-MIDS, we provide an open specification for multimodal MSK imaging, alongside tools for creating compliant datasets. Adherence to our standard will improve harmonization of imaging data across vendors and institutions and permit the development of reproducible processing pipelines and data repositories for MSK research.