PMoC3D: Paired Motion-Corrupted and Motion-Free 3D Brain MRI Scans

PMoC3D is a real-world dataset of Paired Motion-Corrupted 3D brain MRI data. It consists of unprocessed undersampled measurement data for 27 motion-corrupted volumes from 8 subjects, each paired with a motion-free reference scan. The dataset was created by mli-lab and last updated on Hugging Face in November 2025.

Use Cases

• Training motion artifact correction models based on paired corrupted and clean MRI scans.
• Benchmarking image reconstruction algorithms using the provided undersampled measurement data.
• Studying the characteristics of real-world motion artifacts in 3D brain MRI sequences.

Strengths

• Provides 27 motion-corrupted volumes paired with reference scans, enabling supervised learning approaches.
• Includes raw, unprocessed undersampled measurement data, offering flexibility for reconstruction research.
• Data is organized by subject and scan within a sourcedata/ directory.

Limitations

• Dataset size is unknown; the 27 volumes from 8 subjects may be a limited scale for some applications.
• Column-level documentation and file format details are absent, requiring inspection after download.
• Description metadata is limited; actual data quality and specifics require manual verification.

Provenance

Source

mli-lab on Hugging Face

Collection Method

Collection of paired motion-corrupted and motion-free 3D brain MRI scans.

Time Range

null

Freshness

Last updated 2025-11-19 01:55:10

Geography

null