Musculoskeletal (MSK)AI / InformaticsResearch
LumbarSR Dataset Pairs Clinical CT with Micro-CT of 30 Human Lumbar Vertebrae for Super-Resolution
Radiology AI literature (PubMed)Jun 26
A new dataset of 30 human lumbar vertebrae registers clinical CT with micro-CT (105 μm) to help develop super-resolution methods for visualizing trabecular bone microarchitecture, invisible on routine CT.
- Dataset: 30 dry lumbar vertebral specimens, paired micro-CT (105 μm) and clinical CT (8 acquisition configurations: 2 in-plane resolutions, 2 slice thicknesses, 2 kernels).
- Baseline super-resolution benchmarks using deep learning and interpolation are provided, along with image quality metrics and trabecular morphometry.
- Limitation: Specimens are dry teaching specimens without clinical histories, limiting generalizability to in vivo tissue with degenerative changes.
Automated summary
RadPigeon summaries are original and for information only. They are not clinical advice.