Dataset preprocessing
Skill aizech/clinical-skills/.bob/skills/dataset-preprocessing
A collection of AI agent skills focused on medical imaging and healthcare workflows. Built for radiologists, healthcare IT professionals, and researchers who want AI coding agents to help with imaging workflows, clinical documentation, AI integration, and medical research. Works with Claude Code, Codex, Cursor, Windsurf, and many other agents.
npx -y skills add aizech/clinical-skills --skill dataset-preprocessingAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 3 stars3 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Provides preprocessing pipelines and techniques for radiology datasets used in AI development. Use when user mentions "preprocess radiology data", "DICOM preprocessing", "image normalization", "data augmentation", or needs to prepare datasets.
SKILL.md
2.3 KB, as published. Nobody here has run it
Dataset Preprocessing Skill
Triggers
- "preprocess radiology data"
- "DICOM preprocessing"
- "image normalization"
- "data augmentation"
- "quality control pipeline"
- "mask generation"
- "multi-site harmonization"
- "training data preparation"
Parameters
input_format(required): Source data formatdicom- DICOM filesnifti- NIfTI volumesmetadata- Header/excel datamixed- Multiple formats
task_type(required): Downstream ML taskdetection- Object/bounding box detectionsegmentation- Pixel-level segmentationclassification- Image classificationregression- Continuous value prediction
modality(optional): Imaging modalitymulti_vendor(optional): Boolean for multi-site/multi-vendor datadataset_scale(optional): Small (<1K), medium (1K-100K), large (>100K)
Preprocessing Components
Image Processing
- Intensity normalization (z-score, min-max, percentile-based)
- Windowing/leveling for CT/MRI
- Resampling to isotropic voxel size
- Brain extraction (skull stripping)
- Bias field correction for MRI
Quality Control
- Automated quality scoring
- Artifact detection
- Contrast-to-noise ratio
- Resolution verification
- Human-in-the-loop review for edge cases
Augmentation
- Geometric: rotation, flip, scale, elastic deformation
- Intensity: noise, contrast, brightness
- Modality-specific: CT windowing variants, MRI sequence mixing
- Generative: synthetic data augmentation
Format Conversion
- DICOM to NumPy/PyTorch/TensorFlow
- DICOM to NIfTI for volumetric data
- Annotation format conversion (CSV, COCO, YOLO, Pascal VOC)
Output Format
Returns structured JSON with:
- Processing pipeline steps
- Code snippets for each transformation
- Validation checks and statistics
- Expected output specifications
- Common pitfalls and mitigations
Usage Examples
input_format: dicom
task_type: detection
modality: CT
multi_vendor: true
input_format: nifti
task_type: segmentation
dataset_scale: large