Parkinson’s Progression Markers Initiative MRI Scans for 89 Subjects

17,044 DICOM-format MRI images from the Parkinson’s Progression Markers Initiative (PPMI) dataset, contributed by Dhanush Rachaveti on figshare. The collection includes 8,610 images from 47 Parkinson's disease patients and 8,434 images from 42 healthy controls, last updated on 2026-05-23. It contains T1-weighted, T2-weighted, and SWI scans, with some including Diffusion Tensor Imaging (DTI) and functional MRI (fMRI) data for connectivity analysis.

Use Cases

• Parkinson's disease classification based on labeled MRI scans and patient diagnosis metadata
• Biomarker identification for disease progression using detailed scan metadata like voxel size and slice thickness
• Deep learning-based analysis of brain structure changes focusing on regions of interest such as the Substantia Nigra
• Analyzing changes in brain connectivity using included Diffusion Tensor Imaging (DTI) and functional MRI (fMRI) data

Strengths

• 17,044 images with a balanced split of 8,610 PD and 8,434 non-PD images
• Rich DICOM metadata includes Patient Age, Weight, Slice Thickness, Pixel Spacing, and Voxel Size
• Scans target specific regions of interest (ROI) relevant to Parkinson's disease, such as the Substantia Nigra and Basal Ganglia
• Includes multiple MRI modalities (T1, T2, SWI) and some advanced sequences (DTI, fMRI)

Limitations

• Column-level documentation is absent; field semantics must be inferred after download
• Row count is unknown, which may limit suitability assessment
• Data may reflect temporal or recruitment bias inherent to the PPMI study design

Provenance

Source

Parkinson’s Progression Markers Initiative (PPMI)

Collection Method

Clinical MRI scans collected from 47 Parkinson's disease patients and 42 healthy controls.

Freshness

Last updated 2026-05-23 08:36:02; freshness should be verified