Thirty-four patients (mean age: 60.4 ± 15.2 years, age range: 30.0 – 88.1 years, 58.8% females) with the following medical indications for MRI acquisition during the clinical routine were included in this study: 1) lumbar back pain with or without radiculopathy due to suspected spinal degeneration (41.3% of patients), 2) postoperative or follow-up imaging after resection of a spinal tumor (23.5% of patients), 3) known malignancy with (suspected) spinal metastases (17.6% of patients), 4) spondylodiscitis or other spinal inflammatory/infectious diseases (14.7% of patients), and 5) spinal fracture (2.9% of patients). Various scanner models from three different vendors were utilized for collecting a number of images, including: - non-contrast-enhanced T1-weighted, - contrast-enhanced T1-weighted, - T2-weighted, short tau inversion recovery (STIR), - T2-weighted Dixon, - and T1-weighted fat saturation (fs) contrast-enhanced images. 70.9% of included datasets were derived from Philips scanners (Achieva, Ingenia, and Elition), 27.3% were derived from Siemens scanners (Avanto, Verio, Espree, Symphony, Amira, Aera, and Magnetom). One dataset (1.8%) was taken from a GE scanner (Signa). The field of view (FOV) covered at least the entire lumbar spine (L1 to L5). Only scans that were performed in supine position were included. Each patient had at least one MRI scan performed on a Philips system with the following reference sequences: - Sagittal non-contrast-enhanced T1-weighted sequence: repetition time (TR) / echo time (TE) = 600 / 8 ms, FOV = 180 x 275 x 49 mm, acquisition voxel size = 0.80 x 1.00 x 3.00 mm, acquisition duration = 3 min 3 s. - Sagittal T2-weighted DIXON turbo spin-echo (TSE) sequence: TR / TE = 2,500 / 100 ms, FOV = 180 x 275 x 49 mm, acquisition voxel size = 0.70 x 0.98 x 3.00 mm, acquisition duration = 3 min 25 s. The manual segmentations of each vertebral body and intervertebral disc are available as separate binary masks, where pixels with an intensity value of 1 correspond to the tissue of interest, while pixels of an intensity value of 0 belong to the background. Each mask of each image volume was stored as a separate *.nii.gz file. In total, each available volume is accompanied with the corresponding segmentations of 5 vertebral bodies and 4 intervertebral discs. > **Database content and additional files** The database on the OSF consists of 2 components: - **Data** - contains MR images and their respective segmentation maps sorted by patient ID and scanner vendor(s). All images are named per sequence and stored in directory ***Images***, with the accompanying labels stored in ***Labels*** directory, where ***L1-L5*** labels denote lumbar vertebral bodies, while ***BS_L1_2, BS_L2_3*** etc. denote intervertebral discs. - **Patience and sequence metadata** - contains two tables with information about patient characteristics (gender, age) and clinical indications, as well as scan characteristics of each image sequence with information about scanner vendor/model types and scan time intervals. Both tables are saved as .pdf files. The file ***metadata.xlsx*** contains the unified information from both tables. - **Visual examples** - contains example image(s) of segmented lumbar vertebral bodies and intervertebral discs for one patient per all sequences acquired for that patient. > **Reading / visualizing the data** All imaging data, as well as the segmentation maps, are saved using the Neuroimaging Informatics Technology Initiative (NIfTI) format ([https://nifti.nimh.nih.gov/][1]). NIfTI has several features, some of which is encoded in its header, such as: 1. raw data saved in 3D, containing two affine coordinates to relate voxel to spatial index, 2. key acquisition parameters 3. encoding directions 4. grid spacing NIfTI files can be directly read in a number of programming environments, such as **Python**, **Matlab**, and **R**, as well as directly visualized using tools such as **ImageJ** ([https://imagej.nih.gov/ij/][2]), **ITK-Snap** ([http://www.itksnap.org/][3]), **FSL** ([https://fsl.fmrib.ox.ac.uk/fsl/fslwiki][4]), **AFNI** ([https://afni.nimh.nih.gov/][5]) and **FreeSurfer** ([https://surfer.nmr.mgh.harvard.edu/][6]), among others. [1]: https://nifti.nimh.nih.gov/ [2]: https://imagej.nih.gov/ij/ [3]: http://www.itksnap.org/ [4]: https://fsl.fmrib.ox.ac.uk/fsl/fslwiki [5]: https://afni.nimh.nih.gov/ [6]: https://surfer.nmr.mgh.harvard.edu/ We additionally provide 2 small scripts below written in Python using open-source image analysis libraries – **SimpleITK** (https://simpleitk.org/) and **NiBabel** (https://nipy.org/nibabel/) to load images/segmentations for further processing. 1. **NiBabel** We import the nibabel library and load the image and its corresponding L1 segmentation map, where we assume the code is run in the same directory where the image/segmentation files are: import nibabel as nib img=nib.load("T1_contrast.nii.gz") L1_label=nib.load("L1.nii.gz") We can extract the shape, affine transformation information etc. from the loaded image: img_shape=img.shape img_affine=img.affine This information is available without the need to load anything of the main image data into the memory. We can also access the image data as the NumPy array: img_data=img.get_fdata() L1_label_data=L1_label.get_fdata() The complete information embedded in an image header is available through a header object, retrieved as: img_header=img.header L1_label_header=L1_label.header A more detailed tutorial on utilizing the NiBabel API can be found here: https://nipy.org/nibabel/gettingstarted.html. 2. **SimpleITK** Import the SimpleITK Python module and load the image and its corresponding L1 segmentation map objects: import SimpleITK as sitk img=sitk.ReadImage("T1_contrast.nii.gz") L1_label=sitk.ReadImage("L1.nii.gz") Basic image attributes, such as origin, size, spacing and direction can be accessed as: img_origin=img.GetOrigin() img_size=img.GetSize() img_spacing=img.GetSpacing() img_direction=img.GetDirection() Image data can be accessed as a NumPy array as follows: img_data=sitk.GetArrayFromImage(img) L1_label_data=sitk.GetArrayFromimage(L1_label) Detailed SimpleITK documentation and tutorials can be found here: https://simpleitk.org/.