What imaging modalities can I access with this skill?

You can access a wide range of modalities including Computed Tomography (CT), Magnetic Resonance (MR), Positron Emission Tomography (PET), and Slide Microscopy (SM).

Can I preview images before downloading them?

Yes, the skill allows you to generate viewer URLs for specific SeriesInstanceUIDs, which can be opened in a browser for visualization without local software.

How do I filter for specific cancer types?

You can use the client.sql_query() method to join the index table with the collections_index table and filter by the CancerTypes column.

Do I need an API key or account to access IDC data?

No, this skill provides access to public datasets from the NCI Imaging Data Commons that do not require authentication for querying or downloading.

Are the annotations and segmentations also available?

Yes, you can query the analysis_results_index and seg_index tables to find expert-curated or AI-generated segmentations and link them to source imaging data.

Imaging Data Commons

Name: Imaging Data Commons
Author: x-cmd

byx-cmd

•

数据科学与机器学习

Queries and downloads large-scale public cancer imaging datasets from the NCI Imaging Data Commons for AI training and medical research.

This skill leverages the idc-index Python package to provide a streamlined interface for the National Cancer Institute's Imaging Data Commons (IDC). It enables researchers and developers to perform complex SQL-based metadata queries across massive repositories of CT, MR, PET, and pathology images without requiring authentication. By integrating DICOM metadata exploration with direct download capabilities and browser-based visualization, it significantly accelerates the data acquisition workflow for medical AI development and clinical research projects.

主要功能

01Execute SQL queries to filter datasets by modality, cancer type, and anatomical site

02Download DICOM files and series directly for local AI model training

039 GitHub stars

04Access detailed metadata for radiology (CT, MR, PET) and pathology (slide microscopy)

05Generate browser-based viewer URLs for instant medical image visualization

06Verify data licenses and clinical metadata for research compliance

使用场景

01Curating specialized medical imaging datasets for training deep learning models

02Conducting large-scale retrospective studies using public cancer imaging repositories

03Validating AI-generated segmentations and annotations against expert-curated ground truth

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add x-cmd/skill imaging-data-commons

For use in Claude.ai and ChatGPT

Download Skill