SPIDER - Lumbar Spine Segmentation in MR Images: A Dataset and a Public Benchmark

Creators

• Jasper Willem van der Graaf¹
• Miranda L. van Hooff¹
• Constantinus F. M. Buckens¹
• Matthieu Rutten²
• Job L. C. van Susante³
• Robert Jan Kroeze⁴
• Marinus de Kleuver¹
• Bram van Ginneken¹
• Nikolas Lessmann¹

Description

This is a large publicly available multi-center lumbar spine MRI dataset with expert reference segmentations of vertebrae, intervertebral discs (IVDs), and the spinal canal. The dataset contains 447 sagittal T1- and T2-weighted MRI series from 218 studies of 218 patients with a history of low back pain.

Data was collected from four different hospitals. An additional hidden test set is used for evaluation in the accompanying SPIDER challenge at spider.grand-challenge.org.

We release this dataset to promote collaboration and participation in the field of spine segmentation and to improve the diagnostic value of lumbar spine MRI.

Dataset Details

• The overview file includes:
  • Training and validation assignments
  • Biological sex for all patients
  • Patient age (when available)
  • Scanner and acquisition parameters for each MRI study
• A separate file provides radiological gradings per IVD level for the following degenerative changes:
  1. Modic changes (Type I, II, or III)
  2. Endplate changes / Schmorl nodes (binary)
  3. Spondylolisthesis (binary)
  4. Disc herniation (binary)
  5. Disc narrowing (binary)
  6. Disc bulging (binary)
  7. Pfirrman grade (Grade 1 to 5)

Publication

This dataset and the associated public benchmark are described in the following paper:
https://www.nature.com/articles/s41597-024-03090-w

Challenge

The public segmentation challenge associated with this dataset is hosted at:
https://spider.grand-challenge.org/

How to Cite

When using this dataset, please cite both the dataset (with its correct DOI) and the paper linked above.