This readme.txt file was generated on 20240708 by ARIN ELLINGSON Recommended citation for the data: ------------------- GENERAL INFORMATION ------------------- 1. Title of Dataset: Automated MRI-based grading of the lumbar intervertebral disc and facet joints 2. Author Information Principal Investigator Contact Information Name: Arin M Ellingson, PhD Institution: University of Minnesota Address: 420 Delaware St SE; 388 MMC; Minneapolis, MN 55455 Email: ellin224@umn.edu ORCID: https://orcid.org/0000-0001-6154-8035 3. Date published or finalized for release: 25 June 2024 4. Date of data collection (single date, range, approximate date): NA 5. Geographic location of data collection (where was data collected?): NA 6. Information about funding sources that supported the collection of the data: NIH/NCCIH U01-AT010326 and U24-AT011978 7. Overview of the data (abstract): Degeneration of both intervertebral discs (IVDs) and facet joints in the lumbar spine has been associated with low back pain, but whether and how IVD/joint degeneration contributes to pain remains an open question. Joint degeneration can be identified by pairing T1 and T2 MRI with analysis techniques such as Pfirrmann grades (IVD degeneration) and Fujiwara scores (facet degeneration). However, these grades are subjective, prompting the need to develop an automated technique to enhance inter-rater reliability. This study introduces an automated convolutional neural network (CNN) technique trained on clinical MR images of IVD and facet joints obtained from public-access Lumbar Spine MRI Dataset. The primary goal of the automated system is to classify health of lumbar discs and facet joints according to Pfirrmann and Fujiwara grading systems and to enhance inter-rater reliability associated with these grading systems. Performance of the CNN on both the Pfirrmann and Fujiwara scales was measured by comparing the percent agreement and Pearson’s correlation of results from the classifier to the grades assigned by an experienced radiologist. The CNN demonstrates comparable performance to human graders for both Pfirrmann and Fujiwara grading systems, but with larger errors in Fujiwara grading. The CNN improves the reliability of the Pfirrmann system, aligning with previous findings for IVD assessment. The study highlights the potential of using deep learning in classifying the IVD and facet joint health, and due to the high variability in the Fujiwara scoring system highlights the need for improved imaging and scoring techniques to evaluate facet joint health. All codes required to use the automatic grading routines described herein are available in the Data Repository for University of Minnesota (DRUM). -------------------------- SHARING/ACCESS INFORMATION -------------------------- 1. Licenses/restrictions placed on the data: Creative Commons; Attribution, NonCommercial, NoDerivatives 2. Links to publications that cite or use the data: TBD 3. Was data derived from another source? If yes, list source(s): Sudirman S, Al Kafri A, natalia friska, et al. 2019. Lumbar Spine MRI Dataset.2. 4. Terms of Use: Data Repository for the U of Minnesota (DRUM) By using these files, users agree to the Terms of Use. https://conservancy.umn.edu/pages/policies/#drum-terms-of-use --------------------- DATA & FILE OVERVIEW --------------------- 1. File List A. Filename: disc_classification.ipynb Short description: Code to grade intervertebral disc health based on the Pfirrmann scale (MRI) B. Filename: joint_classification.ipynb Short description: Code to grade facet joint health based on the Fujiwara scale (MRI) C. Filename: content.zip Short description: Compressed folder containing the most up to date models for assessing intervertebral disc (disc) and facet joint (joint) health, along with one example image for each. 2. Relationship between files: Disc and Joint Classification files need to access the models embedded in their associated subfolder. -------------------------- METHODOLOGICAL INFORMATION -------------------------- 1. Description of methods used for collection/generation of data: Clinical MR images of lumbar spines were obtained from the public-access Lumbar Spine MRI Dataset (Sudirman S, Al Kafri A, natalia friska, et al. 2019. Lumbar Spine MRI Dataset.2). This dataset consists of anonymized clinical MRI scans of 515 symptomatic patients suffering from back pain. 2. Methods for processing the data: First, the midsagittal and midaxial slices of each motion segment were selected (total of 2633 disc images of motion segments from L1-L2 to L5-S1 and 2377 joint images from both sides of the L1-L2 through L5-S1 motion segments) from T2-weighted sagittal and T1-weighted axial views. Four graders (2 PhD trained spine researchers, 1 neurosurgeon, and 1 musculoskeletal radiologist) graded the IVDs using the Pfirrmann grading system while 5 graders (2 PhD trained spine researchers, 2 neurosurgeons, and 1 musculoskeletal radiologist) graded the facet joints on both sides of the spine using the Fujiwara scoring system. A total of 2366 graded IVD images (90% of total input images) were used to train the automated algorithm, while 267 graded IVD images (10% of total input images) were kept aside and used later to test the accuracy of the automated network solution. For the facet joint, 2135 graded images (90% of total input images) were used to train the CNN, while 242 images (10% of total input images) were used to test the accuracy of the automated network solution. 3. Instrument- or software-specific information needed to interpret the data: Python 4. Standards and calibration information, if appropriate: NA 5. Environmental/experimental conditions: NA 6. Describe any quality-assurance procedures performed on the data: NA 7. People involved with sample collection, processing, analysis and/or submission: Author list included