BIC-MAC Medical Imaging Challenge Dataset

BIC-MAC Challenge Dataset is a collection for synthesizing pseudo-CT images from non-attenuation-corrected PET, whole-body MRI, and 2D topograms. The dataset supports the Big Cross-Modal Attenuation Correction challenge, enabling CT-less PET reconstruction. It was created by DEPICT-RH and last updated in March 2026.

Use Cases

• Train models to synthesize pseudo-CT images from non-attenuation-corrected PET data.
• Develop cross-modal algorithms using whole-body MRI and 2D topograms for attenuation correction.
• Benchmark performance in the BIC-MAC challenge for CT-less PET reconstruction.

Strengths

• Dataset is associated with a formal medical imaging challenge (BIC-MAC).
• Contains multi-modal medical imaging data including PET, MRI, and topograms.
• Last updated in March 2026, indicating recent maintenance.

Limitations

• Dataset size, row count, and specific file formats are unknown.
• Access requires registration at an external website, creating a potential barrier.
• Specific data quality metrics, such as image resolution or patient cohort details, are not provided.

Provenance

Source

DEPICT-RH via Hugging Face.

Collection Method

Created for the BIC-MAC medical imaging challenge.

Freshness

Last updated on 2026-03-23.