Annotated T2-Weighted MR Images of the Lower Spine with 23 Patient Cases

23 anonymized patients' T2-weighted turbo spin echo MR spine images, each containing at least 7 vertebral bodies from T11 to L5. The dataset includes reference manual segmentations provided as binary masks for each vertebral body, stored in the Neuroimaging Informatics Technology Initiative (NIFTI) file format. This resource was made available by an international consortium of universities and research institutes for developing spinal image analysis algorithms.

Use Cases

• Training segmentation models for vertebral bodies based on the provided binary mask annotations.
• Developing and benchmarking 3D medical image analysis tools using the NIFTI format spine images.
• Researching automated localization methods for lower spine vertebrae (T11-L5) as referenced in the associated publication.

Strengths

• Includes reference manual segmentations as binary masks for each vertebral body.
• Data from 23 patients provides a basis for algorithm development.
• Images cover at least 7 vertebral bodies per patient, offering consistent anatomical scope.

Limitations

• Row count and dataset size are unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Annotations were prepared by a single individual without professional radiology training, as noted in the description.

Provenance

Source

Institute for Surgical Technology and Biomechanics, University of Bern, Switzerland, and collaborating institutions.

Collection Method

T2-weighted turbo spin echo MR imaging of patient spines.

Geography

Data collection involved institutions in Switzerland, Germany, and Australia.