OmniLabelX
Intelligent Medical Image Annotation
AI-powered annotation platform that reduces labeling time by 70%. Support for 2D/3D segmentation, multi-modal imaging, and collaborative workflows with DICOM-SR compliant exports.
Everything You Need to Annotate Medical Images
From simple bounding boxes to complex 3D volumetric segmentation
2D Annotation Tools
Comprehensive toolkit including bounding boxes, polygons, freehand drawing, brush/eraser, and smart scissors for precise delineation.
3D Volumetric Segmentation
MONAI-powered 3D segmentation for CT and MRI volumes. Slice-by-slice or automated propagation with manual refinement.
AI Pre-labeling
YOLOv8 and SAM (Segment Anything Model) generate initial annotations. Reduce manual work by up to 70%.
Multi-user Collaboration
Real-time collaboration with role-based access. Annotators, reviewers, and approvers work in a structured workflow.
Quality Assurance
Built-in QA with inter-annotator agreement metrics, Dice coefficient tracking, and gold standard validation.
Standard Exports
Export to COCO JSON, NIfTI, DICOM-SR, and Pascal VOC. Full FHIR compatibility for clinical integration.
How OmniLabelX Works
Upload
Import DICOM from PACS or upload directly. Automatic metadata extraction.
Pre-label
AI models generate initial annotations. Review and adjust as needed.
Refine
Use annotation tools to perfect labels. Collaborate with team members.
Export
QA validation, approval workflow, and export in your preferred format.
Supported Imaging Modalities
Intuitive, Powerful Interface
Built with Fabric.js and Cornerstone.js, OmniLabelX provides a responsive, feature-rich annotation canvas that handles everything from simple X-rays to complex multi-frame CT volumes.
- ✓ Pan, zoom, and window/level controls
- ✓ Multi-planar reconstruction (MPR)
- ✓ Measurement tools (length, area, HU)
- ✓ Keyboard shortcuts for efficiency
- ✓ Undo/redo with full history
Technical Details
| Specification | Details |
|---|---|
| Supported Formats | DICOM, NIfTI, PNG, JPEG, TIFF, SVS (pathology) |
| Export Formats | COCO JSON, Pascal VOC, NIfTI, DICOM-SR, FHIR Observation |
| AI Models | YOLOv8, SAM (Segment Anything), MONAI UNet, Custom fine-tuned models |
| Annotation Types | Bounding box, polygon, polyline, freehand, brush, 3D mask, landmarks |
| Max Image Size | Up to 100,000 x 100,000 pixels (pathology), unlimited DICOM series |
| Collaboration | Real-time multi-user, role-based access, audit trail |
| Standards | DICOM, DICOMweb, FHIR R5, SNOMED-CT, ICD-10 |
| Browser Support | Chrome 90+, Firefox 88+, Safari 14+, Edge 90+ |
Ready to Accelerate Your Annotation Workflow?
See how OmniLabelX can reduce your annotation time by 70% while improving quality and consistency.