Skip to content

Conversation

Copy link

Copilot AI commented Sep 17, 2025

This PR adds comprehensive VISTA3D tutorials and documentation to MONAI Label, addressing the need for enhanced guidance on using the VISTA3D universal segmentation model.

What's Added

Comprehensive Tutorial Documentation

  • Complete VISTA3D Tutorial (docs/source/tutorials/vista3d_tutorial.md): 270+ line comprehensive guide covering setup, usage, troubleshooting, and clinical workflows
  • Quick Start Guide (docs/source/tutorials/vista3d_quickstart.md): Fast setup guide for immediate VISTA3D deployment
  • Configuration Examples (docs/source/tutorials/vista3d_configurations.md): Sample configurations for various deployment scenarios including memory optimization, PACS integration, and Docker deployment

Main README Enhancements

  • Added dedicated VISTA3D App section to the tutorials listing
  • Updated the supported models matrix to include VISTA3D (Universal) in the Radiology section
  • Linked all three VISTA3D tutorial documents with clear descriptions

MONAIBundle App Integration

  • Added VISTA3D to the supported models table in sample-apps/monaibundle/README.md
  • Included VISTA3D-specific usage examples and command-line configurations
  • Added new configuration parameters for VISTA3D tuning

Formal Documentation System

  • Created docs/source/tutorials.rst for Sphinx documentation integration
  • Added tutorials section to the main documentation index (docs/source/index.rst)
  • Provided proper RST formatting for the documentation system

Key Features Covered

Setup and Installation:

  • System requirements (GPU, RAM, Python versions)
  • Installation methods including pip and development setups
  • Data preparation for CT volumes (DICOM and NIfTI)

Usage Workflows:

  • Automatic segmentation for 117+ anatomical structures
  • Interactive point prompting for targeted segmentation
  • Class-based prompting for anatomical categories (organs, bones, vessels)
  • Advanced configuration and memory optimization

Viewer Integration:

  • OHIF viewer setup and VISTA3D integration
  • 3D Slicer compatibility and usage instructions
  • Web-based and desktop workflow examples

Deployment Scenarios:

  • Local datastore configurations
  • PACS/DICOMWeb integration
  • Docker deployment examples
  • Multi-model server setups

Clinical Applications:

  • Abdominal organ segmentation workflows
  • Comprehensive anatomy analysis
  • Research and clinical use case examples

The tutorials provide both quick-start capabilities for immediate use and comprehensive guidance for advanced deployment scenarios, making VISTA3D's universal segmentation capabilities accessible to users across different experience levels.

Fixes #1859.

Warning

Firewall rules blocked me from connecting to one or more addresses (expand for details)

I tried to connect to the following addresses, but was blocked by firewall rules:

  • https://api.github.com/repos/Project-MONAI/model-zoo/contents/models
    • Triggering command: python -c import requests try: response = requests.get('REDACTED') if response.status_code == 200: models = response.json() vista_models = [model['name'] for model in models if 'vista' in model['name'].lower()] print('VISTA models in Model Zoo:', vista_models) else: print('Could not fetch model zoo contents') except Exception as e: print('Error fetching model zoo:', str(e)) (http block)
    • Triggering command: `python -c
      import requests
      try:
      # Check the public model zoo for VISTA3D models
      url = 'REDACTED'
      response = requests.get(url, timeout=10)
      if response.status_code == 200:
      print('VISTA3D found in model zoo')
      print(response.text[:200])
      else:
      print('VISTA3D not found at expected location')
# Try alternative approaches to find VISTA3D
models_list_url = 'REDACTED'
response = requests.get(models_list_url, timeout=10)
if response.status_code == 200:
    import json
    models = json.loads(response.text)
    vista_models = [m['name'] for m in models if 'vista' in m['name'].lower()]
    print('Available VISTA models in model zoo:', vista_models)
else:
    print('Could not retrieve model list')

except Exception as e:
print(f'Error: {e}')` (http block)

If you need me to access, download, or install something from one of these locations, you can either:


💬 Share your feedback on Copilot coding agent for the chance to win a $200 gift card! Click here to start the survey.

Copy link
Author

Copilot AI commented Sep 17, 2025

@bhashemian 👋 This repository doesn't have Copilot instructions. With Copilot instructions, I can understand the repository better, work faster and produce higher quality PRs.

I can generate a .github/copilot-instructions.md file for you automatically. Click here to open a pre-filled issue and assign it to me. I'll write the instructions, and then tag you for review.

Copilot AI changed the title [WIP] Enhance MONAI label tutorial on using VISTA3D Enhance MONAI Label tutorial on using VISTA3D Sep 17, 2025
Copilot AI requested a review from bhashemian September 17, 2025 14:46
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Enhance MONAI label tutorial on using VISTA3D

2 participants